技术领域Technical field
本发明属于智能交通相关技术领域，更具体地，涉及一种运输车辆穿越道路时的交通预警方法及系统。The present invention belongs to the technical field related to intelligent transportation, and more specifically, relates to a traffic warning method and system for transport vehicles crossing roads.
背景技术Background technique
传统的交通安全监测方法主要依赖于人工巡逻和交通标志提示，但这些方法效率低下且主观性强，无法实现对行人和车辆状态的准确监测和预警。此外，由于乡镇环境道路大多没有较为全面的监测设备，道路设置较为简单，通常没有正规的多种类型行车道设置且道路两边往往存在有住宅或商店导致道路两边行人活动较为复杂，由于乡镇环境道路的复杂性，现有的监测技术在乡镇道路上的应用存在困难。另外，由于施工现场的特殊性，施工物料的运输车辆等大型货运车辆行驶在道路上时危险系数相比一般车辆更高，尤其是对于乡镇道路上的运输车辆，对其进行交通监测预警是非常有必要的。Traditional traffic safety monitoring methods mainly rely on manual patrols and traffic sign prompts, but these methods are inefficient and highly subjective, and cannot achieve accurate monitoring and early warning of the status of pedestrians and vehicles. In addition, because most of the township environmental roads do not have comprehensive monitoring equipment, the road settings are relatively simple, there are usually no formal multi-type lane settings, and there are often residences or shops on both sides of the road, which makes pedestrian activities on both sides of the road more complicated. Due to the township environmental roads Due to the complexity, it is difficult to apply existing monitoring technology on township roads. In addition, due to the particularity of the construction site, large freight vehicles such as construction material transport vehicles have a higher risk factor when driving on the road than ordinary vehicles. Especially for transport vehicles on township roads, it is very important to conduct traffic monitoring and early warning. Necessary.
近年来，计算机视觉技术在交通安全领域得到了广泛应用。通过使用摄像头和传感器等设备，可以获取道路上车辆和行人的图像和数据。基于计算机视觉的交通安全监测方法还能够结合其他传感器数据，如雷达和激光扫描仪等，提高监测的准确性和可靠性。In recent years, computer vision technology has been widely used in the field of traffic safety. By using devices such as cameras and sensors, images and data of vehicles and pedestrians on the road can be acquired. Traffic safety monitoring methods based on computer vision can also be combined with other sensor data, such as radar and laser scanners, to improve the accuracy and reliability of monitoring.
然而，目前存在的基于计算机视觉的交通安全监测方法主要集中在道路交通流量、车辆违规行为等方面，针对运输车辆穿越道路时的交通安全监测和提醒预警的研究还相对较少。因此，需要提出一种新的方法来解决该问题，以实现对运输车辆的交通安全预警。However, the currently existing traffic safety monitoring methods based on computer vision mainly focus on road traffic flow, vehicle violations, etc., and there are relatively few studies on traffic safety monitoring and warnings when transport vehicles cross roads. Therefore, a new method needs to be proposed to solve this problem to achieve traffic safety warning for transport vehicles.
发明内容Contents of the invention
针对现有技术的以上缺陷或改进需求，本发明提供了一种运输车辆穿越道路时的交通预警方法及系统，解决了目前存在的基于计算机视觉的交通安全监测方法缺乏针对运输车辆穿越道路时的交通安全监测和提醒预警的问题，提出了一种针对运输车辆穿越道路时的交通预警方案，实现对运输车辆的准确监测、区域安全判定和安全状况预警。In view of the above defects or improvement needs of the existing technology, the present invention provides a traffic warning method and system for transport vehicles crossing roads, which solves the lack of current computer vision-based traffic safety monitoring methods for transport vehicles crossing roads. Regarding the issue of traffic safety monitoring and reminder and early warning, a traffic early warning scheme for transport vehicles crossing roads is proposed to achieve accurate monitoring of transport vehicles, regional safety determination and safety status early warning.
为实现上述目的，按照本发明的一个方面，提供了一种运输车辆穿越道路时的交通预警方法，包括：In order to achieve the above object, according to one aspect of the present invention, a traffic warning method for transport vehicles crossing roads is provided, including:
对道路状态进行监测，获取道路监控信息；Monitor road status and obtain road monitoring information;
根据所述道路监控信息，获取道路上的运输车辆以及行人相关信息，其中道路上的运输车辆以及行人相关信息包括所述运输车辆的数量和位置信息、所述行人的数量和位置信息、所述行人的人员组成信息以及所述运输车辆和所述行人的轨迹入侵预测信息；According to the road monitoring information, information related to transportation vehicles and pedestrians on the road is obtained, where the information related to transportation vehicles and pedestrians on the road includes the number and location information of the transportation vehicles, the number and location information of the pedestrians, the Pedestrian composition information and trajectory intrusion prediction information of the transportation vehicle and the pedestrian;
根据道路上的运输车辆以及行人相关信息判断道路安全等级，其中所述道路安全等级预先设置为包括第一安全等级、第二安全等级和第三安全等级三个级别；Determine the road safety level based on information related to transportation vehicles and pedestrians on the road, where the road safety level is preset to include three levels: a first safety level, a second safety level and a third safety level;
根据所述道路安全等级对所述运输车辆和/或所述行人进行预警提醒。Provide an early warning to the transport vehicle and/or the pedestrian according to the road safety level.
根据本发明提供的运输车辆穿越道路时的交通预警方法，所述行人的人员组成信息具体包括：According to the traffic warning method when transport vehicles cross roads provided by the present invention, the pedestrian composition information specifically includes:
所述行人中成年人的数量和未成年人的数量。The number of adults and the number of minors among said pedestrians.
根据本发明提供的运输车辆穿越道路时的交通预警方法，获取所述行人的人员组成信息具体包括：According to the traffic warning method when transport vehicles cross roads provided by the present invention, obtaining the pedestrian composition information specifically includes:
根据所述道路监控信息，分析获取所述行人的尺寸大小；Analyze and obtain the size of the pedestrian according to the road monitoring information;
在所述行人的尺寸大小小于预设值时，判定所述行人是未成年人。When the size of the pedestrian is smaller than a preset value, it is determined that the pedestrian is a minor.
根据本发明提供的运输车辆穿越道路时的交通预警方法，获取所述运输车辆和所述行人的轨迹入侵预测信息具体包括：According to the traffic warning method when a transport vehicle crosses a road provided by the present invention, obtaining the trajectory intrusion prediction information of the transport vehicle and the pedestrian specifically includes:
根据所述道路监控信息，分析获取所述运输车辆和所述行人分别的轨迹预测信息；Analyze and obtain the trajectory prediction information of the transportation vehicle and the pedestrian according to the road monitoring information;
根据所述运输车辆和所述行人分别的轨迹预测信息，分析判断所述行人的轨迹入侵所述运输车辆轨迹的可能性。According to the respective trajectory prediction information of the transportation vehicle and the pedestrian, the possibility of the pedestrian's trajectory invading the transportation vehicle trajectory is analyzed and determined.
根据本发明提供的运输车辆穿越道路时的交通预警方法，以下情形下判定所述道路安全等级为第一安全等级：According to the traffic warning method for transport vehicles crossing roads provided by the present invention, the road safety level is determined to be the first safety level under the following circumstances:
道路上无所述运输车辆；There are no such transport vehicles on the road;
或者道路上存在所述运输车辆，所述行人的总数量小于A，且不存在未成年人；Or there are said transport vehicles on the road, the total number of said pedestrians is less than A, and there are no minors;
或者道路上存在所述运输车辆，所述行人中未成年人的数量小于B，且所述行人的轨迹预测不侵入所述运输车辆轨迹；Or the transport vehicle exists on the road, the number of minors among the pedestrians is less than B, and the trajectory prediction of the pedestrians does not invade the trajectory of the transport vehicle;
其中，A大于B。Among them, A is greater than B.
根据本发明提供的运输车辆穿越道路时的交通预警方法，以下情形下判定所述道路安全等级为第二安全等级：According to the traffic warning method for transport vehicles crossing roads provided by the present invention, the road safety level is determined to be the second safety level under the following circumstances:
道路上存在所述运输车辆，所述行人的总数量大于等于A、小于C，且不存在未成年人；The transport vehicle exists on the road, the total number of pedestrians is greater than or equal to A but less than C, and there are no minors;
或者道路上存在所述运输车辆，所述行人的总数量小于A，且所述行人中未成年人的数量大于等于B、小于D；Or the transport vehicle exists on the road, the total number of pedestrians is less than A, and the number of minors among the pedestrians is greater than or equal to B but less than D;
或者道路上存在所述运输车辆，所述行人中未成年人的数量小于B，且至少一个所述行人的轨迹预测侵入所述运输车辆轨迹；Or there is the transport vehicle on the road, the number of minors among the pedestrians is less than B, and the trajectory of at least one of the pedestrians is predicted to invade the trajectory of the transport vehicle;
其中，C、A、D、B的数量依次减小。Among them, the numbers of C, A, D, and B decrease in order.
根据本发明提供的运输车辆穿越道路时的交通预警方法，以下情形下判定所述道路安全等级为第三安全等级：According to the traffic warning method when transport vehicles cross roads provided by the present invention, the road safety level is determined to be the third safety level under the following circumstances:
道路上存在所述运输车辆，所述行人的总数量大于等于C，且不存在未成年人；The transport vehicle exists on the road, the total number of pedestrians is greater than or equal to C, and there are no minors;
或者道路上存在所述运输车辆，所述行人中未成年人的数量大于等于D，且所述行人的轨迹预测不侵入所述运输车辆轨迹；Or the transport vehicle exists on the road, the number of minors among the pedestrians is greater than or equal to D, and the trajectory prediction of the pedestrians does not invade the trajectory of the transport vehicle;
或者道路上存在所述运输车辆，所述行人的总数量小于C，所述行人中未成年人的数量大于等于B、小于DA，且至少一个所述行人的轨迹预测侵入所述运输车辆轨迹；Or the transport vehicle exists on the road, the total number of pedestrians is less than C, the number of minors among the pedestrians is greater than or equal to B and less than DA, and the trajectory of at least one of the pedestrians is predicted to invade the trajectory of the transport vehicle;
其中，C、D、B的数量依次减小。Among them, the numbers of C, D, and B decrease in order.
根据本发明提供的运输车辆穿越道路时的交通预警方法，当道路安全等级为第一安全等级时，不进行预警提醒；当道路安全等级为第二安全等级时，提醒所述运输车辆减速通过，且提醒所述行人远离道路；当道路安全等级为第三安全等级时，提醒所述运输车辆暂缓通行，提醒所述行人远离道路。According to the traffic warning method when a transport vehicle crosses a road provided by the present invention, when the road safety level is the first safety level, no warning reminder is provided; when the road safety level is the second safety level, the transport vehicle is reminded to slow down and pass. And remind the pedestrians to stay away from the road; when the road safety level is the third safety level, remind the transportation vehicles to temporarily stop passing, and remind the pedestrians to stay away from the road.
按照本发明的另一个方面，提供了一种运输车辆穿越道路时的交通预警系统，包括：According to another aspect of the present invention, a traffic warning system for transport vehicles crossing roads is provided, including:
监测模块，用于对道路状态进行监测，获取道路监控信息；Monitoring module, used to monitor road status and obtain road monitoring information;
分析处理模块，用于根据所述道路监控信息，获取道路上的运输车辆以及行人相关信息，其中道路上的运输车辆以及行人相关信息包括所述运输车辆的数量和位置信息、所述行人的数量和位置信息、所述行人的人员组成信息以及所述运输车辆和所述行人的轨迹入侵预测信息；An analysis and processing module configured to obtain information related to transportation vehicles and pedestrians on the road based on the road monitoring information, where the information related to transportation vehicles and pedestrians on the road includes the number and location information of the transportation vehicles and the number of pedestrians. and location information, the pedestrian composition information, and the trajectory intrusion prediction information of the transportation vehicle and the pedestrian;
安全判定模块，用于根据道路上的运输车辆以及行人相关信息判断道路安全等级，其中所述道路安全等级预先设置为包括第一安全等级、第二安全等级和第三安全等级三个级别；A safety determination module, used to determine the road safety level based on information related to transport vehicles and pedestrians on the road, where the road safety level is preset to include three levels: a first safety level, a second safety level and a third safety level;
预警模块，用于根据所述道路安全等级对所述运输车辆和/或所述行人进行预警提醒。An early warning module is used to provide early warning to the transportation vehicle and/or the pedestrian according to the road safety level.
根据本发明提供的运输车辆穿越道路时的交通预警系统，所述分析处理模块具体包括：According to the traffic warning system for transport vehicles crossing roads provided by the present invention, the analysis and processing module specifically includes:
目标检测模块，用于根据所述道路监控信息，获取所述运输车辆的数量和位置信息、所述行人的数量和位置信息以及所述行人的人员组成信息；A target detection module, configured to obtain the number and location information of the transport vehicles, the number and location information of the pedestrians, and the personnel composition information of the pedestrians based on the road monitoring information;
时序分析模块，用于根据所述道路监控信息，获取所述运输车辆和所述行人的轨迹入侵预测信息。A time series analysis module, configured to obtain trajectory intrusion prediction information of the transportation vehicle and the pedestrian based on the road monitoring information.
总体而言，通过本发明所构思的以上技术方案与现有技术相比，本发明提供的运输车辆穿越道路时的交通预警方法及系统：Generally speaking, compared with the existing technology through the above technical solutions conceived by the present invention, the traffic warning method and system provided by the present invention when transport vehicles cross roads:
1.通过获取道路监控信息结合视觉识别技术分析道路上的运输车辆以及行人相关信息，然后基于该信息建立多级道路安全等级，在判定道路所属安全等级之后进行交通预警提醒，能够实现对运输车辆的准确监测、区域安全判定和安全状况预警，从而降低道路事故发生率，提高道路安全性；且道路安全等级判定中囊括了多类信息，相比传统的人工观察监控法具有信息获取全面、状况判断准确、决策建议及时的优点；1. By obtaining road monitoring information and combining it with visual recognition technology to analyze information about transportation vehicles and pedestrians on the road, and then establishing multi-level road safety levels based on this information, and then carrying out traffic warning reminders after determining the safety level of the road, it is possible to realize the monitoring of transportation vehicles Accurate monitoring, regional safety determination and safety status early warning, thereby reducing the incidence of road accidents and improving road safety; and the determination of road safety levels includes multiple types of information, compared with the traditional manual observation and monitoring method, it has comprehensive information acquisition, status The advantages of accurate judgment and timely decision-making suggestions;
2.提出对行人中的成年人和未成年人进行区分，进而分析行人中的人员组成，从而在道路安全等级划分时，可重点考虑未成年人的数量以及活动轨迹，使得对道路的安全预警更有针对性，更加实用有效；2. It is proposed to distinguish between adults and minors among pedestrians, and then analyze the composition of pedestrians, so that when classifying road safety levels, the number and activity trajectories of minors can be mainly considered, so as to provide early warning for road safety. More targeted, more practical and effective;
3.道路安全等级的设置通过预设的具体标准，结合自动化的识别技术能够实现检测、分析、预警全流程的智能化、标准化，提高了物料运输车穿越城镇的安全性，进一步地保障了工程的顺利进行。3. The setting of road safety levels can realize the intelligentization and standardization of the entire process of detection, analysis, and early warning through preset specific standards and automated identification technology, which improves the safety of material transport vehicles passing through towns and further guarantees the safety of the project. progressed smoothly.
附图说明Description of drawings
图1是本发明提供的运输车辆穿越道路时的交通预警方法的示意图；Figure 1 is a schematic diagram of a traffic warning method when a transport vehicle crosses a road provided by the present invention;
图2是本发明提供的运输车辆穿越道路时的交通预警系统的示意图。Figure 2 is a schematic diagram of a traffic warning system when a transport vehicle crosses a road provided by the present invention.
具体实施方式Detailed ways
为了使本发明的目的、技术方案及优点更加清楚明白，以下结合附图及实施例，对本发明进行进一步详细说明。应当理解，此处所描述的具体实施例仅仅用以解释本发明，并不用于限定本发明。此外，下面所描述的本发明各个实施方式中所涉及到的技术特征只要彼此之间未构成冲突就可以相互组合。In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.
请参阅图1，本发明提供一种运输车辆穿越道路时的交通预警方法，该交通预警方法包括：Please refer to Figure 1. The present invention provides a traffic warning method when a transport vehicle crosses a road. The traffic warning method includes:
S1，对道路状态进行监测，获取道路监控信息；S1, monitor the road status and obtain road monitoring information;
S2，根据所述道路监控信息，获取道路上的运输车辆以及行人相关信息，其中道路上的运输车辆以及行人相关信息包括所述运输车辆的数量和位置信息、所述行人的数量和位置信息、所述行人的人员组成信息以及所述运输车辆和所述行人的轨迹入侵预测信息；S2, according to the road monitoring information, obtain information related to transportation vehicles and pedestrians on the road, where the information related to transportation vehicles and pedestrians on the road includes the number and location information of the transportation vehicles, the number and location information of the pedestrians, The pedestrian composition information and the trajectory intrusion prediction information of the transportation vehicle and the pedestrian;
S3，根据道路上的运输车辆以及行人相关信息判断道路安全等级，其中所述道路安全等级预先设置为包括第一安全等级、第二安全等级和第三安全等级三个级别；S3, determine the road safety level based on information related to transport vehicles and pedestrians on the road, where the road safety level is preset to include three levels: the first safety level, the second safety level and the third safety level;
S4，根据所述道路安全等级对所述运输车辆和/或所述行人进行预警提醒。S4: Provide an early warning to the transportation vehicle and/or the pedestrian according to the road safety level.
本实施例中获取的道路监控信息可为道路的图像或视频信息。对道路的图像或视频信息进行特征识别，可识别出运输车辆和行人的相关信息。本实施例提出具体根据道路监控信息分析获取运输车辆的数量即有无运输车辆的存在、行人的数量、行人的人员组成即对行人人员进行划分，然后根据运输车辆的位置变化以及行人的位置变化分析运输车辆和行人的轨迹入侵相关信息。The road monitoring information obtained in this embodiment may be image or video information of the road. Feature recognition of road images or video information can identify relevant information about transportation vehicles and pedestrians. This embodiment proposes to specifically obtain the number of transport vehicles based on the analysis of road monitoring information, that is, the presence or absence of transport vehicles, the number of pedestrians, and the composition of pedestrians, that is, to divide the pedestrians, and then based on the location changes of the transport vehicles and the location changes of the pedestrians. Analyze information related to trajectory intrusion of transportation vehicles and pedestrians.
进一步基于上述各信息建立道路安全等级划分的三个预设级别，然后根据实际监控信息判断实际的道路安全等级，针对道路安全等级的判定对道路交通安全进行预警提醒，从而降低道路事故发生率，提高道路安全性。Furthermore, three preset levels of road safety classification are established based on the above information, and then the actual road safety level is determined based on the actual monitoring information. Based on the determination of the road safety level, early warnings and reminders for road traffic safety are provided, thereby reducing the incidence of road accidents. Improve road safety.
本发明提供的运输车辆穿越道路时的交通预警方法，通过获取道路监控信息结合视觉识别技术分析道路上的运输车辆以及行人相关信息，然后基于该信息建立多级道路安全等级，在判定道路所属安全等级之后进行交通预警提醒，能够实现对运输车辆的准确监测、区域安全判定和安全状况预警，从而降低道路事故发生率，提高道路安全性；且道路安全等级判定中囊括了多类信息，相比传统的人工观察监控法具有信息获取全面、状况判断准确、决策建议及时的优点。The traffic warning method provided by the present invention when transport vehicles cross the road is to obtain road monitoring information combined with visual recognition technology to analyze transport vehicles and pedestrian-related information on the road, and then establish a multi-level road safety level based on this information, and then determine the safety of the road. Traffic early warning reminders after the level can achieve accurate monitoring of transport vehicles, regional safety determination and safety status early warning, thereby reducing the incidence of road accidents and improving road safety; and the road safety level determination includes many types of information, compared with The traditional manual observation and monitoring method has the advantages of comprehensive information acquisition, accurate situation judgment, and timely decision-making and suggestions.
进一步地，本实施例中道路可为乡镇道路；即对于复杂性较高的乡镇道路来说，运输车辆在乡镇道路上行驶穿越时对周围环境及行人会造成更高的危险性，因此对于乡镇道路很有必要对运输车辆的穿越进行监控预警。乡镇道路可为省道或规格小于省道的道路；乡镇道路通常为无人行道的道路，且道路两边通常行人活动较为复杂，行人较易与车辆产生干涉。运输车辆可为施工物料运输车辆等大型货运车辆。Furthermore, the roads in this embodiment may be township roads; that is, for township roads with high complexity, transportation vehicles will cause higher risks to the surrounding environment and pedestrians when traveling on township roads, so for township roads It is necessary to monitor and provide early warning for transportation vehicles crossing the road. Township roads can be provincial roads or roads smaller than provincial roads; township roads are usually roads without sidewalks, and pedestrian activities on both sides of the road are usually more complex, making it easier for pedestrians to interfere with vehicles. The transport vehicle may be a large freight vehicle such as a construction material transport vehicle.
进一步地，所述行人的人员组成信息具体包括：Further, the pedestrian composition information specifically includes:
所述行人中成年人的数量和未成年人的数量。本实施例提出对行人中的成年人和未成年人进行区分，进而分析行人中的人员组成，充分考虑了未成年人自身防范意识相对较差从而更易受到伤害的因素，从而在道路安全等级划分时，可重点考虑未成年人的数量以及活动轨迹，使得对道路的安全预警更有针对性，更加实用有效。The number of adults and the number of minors among said pedestrians. This embodiment proposes to distinguish between adults and minors among pedestrians, and then analyze the composition of pedestrians, fully considering the factors that minors have relatively poor awareness of prevention and are more vulnerable to injury, so as to classify road safety levels. When doing so, we can focus on the number and activity trajectories of minors, making road safety warnings more targeted, practical and effective.
具体地，获取所述行人的人员组成信息具体包括：Specifically, obtaining the pedestrian composition information specifically includes:
根据所述道路监控信息，分析获取所述行人的尺寸大小；Analyze and obtain the size of the pedestrian according to the road monitoring information;
在所述行人的尺寸大小小于预设值时，判定所述行人是未成年人。When the size of the pedestrian is smaller than a preset value, it is determined that the pedestrian is a minor.
即对于一定距离下拍摄的图像，成年人会对应一个尺寸大小，未成年人会对应另一个尺寸大小，可预先根据拍摄图像的距离确定预设值，该预设值可根据拍摄距离下成年人应该具有的尺寸大小来设置，从而在对图像中的行人进行分析时，如果尺寸小于该预设尺寸，可判定行人是未成年人，其余的则判定为成年人。具体地，尺寸可为高度尺寸。That is to say, for images taken at a certain distance, adults will correspond to one size, and minors will correspond to another size. The preset value can be determined in advance according to the distance from which the image is taken. The preset value can be determined according to the shooting distance for adults. The size should be set so that when analyzing pedestrians in the image, if the size is smaller than the preset size, the pedestrian can be determined to be a minor, and the rest can be determined to be an adult. Specifically, the dimension may be a height dimension.
进一步地，获取所述运输车辆和所述行人的轨迹入侵预测信息具体包括：Further, obtaining the trajectory intrusion prediction information of the transportation vehicle and the pedestrian specifically includes:
根据所述道路监控信息，分析获取所述运输车辆和所述行人分别的轨迹预测信息；Analyze and obtain the trajectory prediction information of the transportation vehicle and the pedestrian according to the road monitoring information;
根据所述运输车辆和所述行人分别的轨迹预测信息，分析判断所述行人的轨迹入侵所述运输车辆轨迹的可能性。即根据道路监控信息，对运输车辆和行人的移动轨迹进行短时预测，判断行人是否存在入侵至运输车辆轨迹的可能性。具体可根据道路监控信息，例如监控视频，分析每一帧图像中行人的位置，进而形成一个位置时序，可获取行人的移动速度以及对行人下一时刻的移动轨迹进行短时预测。进而分析行人在较短时间内是否存在入侵至运输车辆轨迹中的可能性。According to the respective trajectory prediction information of the transportation vehicle and the pedestrian, the possibility of the pedestrian's trajectory invading the transportation vehicle trajectory is analyzed and determined. That is, based on road monitoring information, short-term predictions are made of the movement trajectories of transport vehicles and pedestrians to determine whether there is a possibility of pedestrians invading the trajectories of transport vehicles. Specifically, based on road monitoring information, such as surveillance video, the position of the pedestrian in each frame of the image can be analyzed to form a position sequence, which can obtain the pedestrian's moving speed and make short-term predictions of the pedestrian's movement trajectory at the next moment. Then analyze whether there is a possibility of pedestrians intruding into the trajectory of transport vehicles in a short period of time.
进一步地，以下情形下判定所述道路安全等级为第一安全等级：Further, the road safety level is determined to be the first safety level under the following circumstances:
道路上无所述运输车辆；There are no such transport vehicles on the road;
或者道路上存在所述运输车辆，所述行人的总数量小于A，且不存在未成年人；Or there are said transport vehicles on the road, the total number of said pedestrians is less than A, and there are no minors;
或者道路上存在所述运输车辆，所述行人中未成年人的数量小于B，且所述行人的轨迹预测不侵入所述运输车辆轨迹；Or the transport vehicle exists on the road, the number of minors among the pedestrians is less than B, and the trajectory prediction of the pedestrians does not invade the trajectory of the transport vehicle;
其中，A大于B。Among them, A is greater than B.
以下情形下判定所述道路安全等级为第二安全等级：The road safety level is determined to be the second safety level under the following circumstances:
道路上存在所述运输车辆，所述行人的总数量大于等于A、小于C，且不存在未成年人；The transport vehicle exists on the road, the total number of pedestrians is greater than or equal to A but less than C, and there are no minors;
或者道路上存在所述运输车辆，所述行人的总数量小于A，且所述行人中未成年人的数量大于等于B、小于D；Or the transport vehicle exists on the road, the total number of pedestrians is less than A, and the number of minors among the pedestrians is greater than or equal to B but less than D;
或者道路上存在所述运输车辆，所述行人中未成年人的数量小于B，且至少一个所述行人的轨迹预测侵入所述运输车辆轨迹；Or there is the transport vehicle on the road, the number of minors among the pedestrians is less than B, and the trajectory of at least one of the pedestrians is predicted to invade the trajectory of the transport vehicle;
其中，C、A、D、B的数量依次减小。Among them, the numbers of C, A, D, and B decrease in order.
以下情形下判定所述道路安全等级为第三安全等级：The road safety level is determined to be the third safety level under the following circumstances:
道路上存在所述运输车辆，所述行人的总数量大于等于C，且不存在未成年人；The transport vehicle exists on the road, the total number of pedestrians is greater than or equal to C, and there are no minors;
或者道路上存在所述运输车辆，所述行人中未成年人的数量大于等于D，且所述行人的轨迹预测不侵入所述运输车辆轨迹；Or the transport vehicle exists on the road, the number of minors among the pedestrians is greater than or equal to D, and the trajectory prediction of the pedestrians does not invade the trajectory of the transport vehicle;
或者道路上存在所述运输车辆，所述行人的总数量小于C，所述行人中未成年人的数量大于等于B、小于D，且至少一个所述行人的轨迹预测侵入所述运输车辆轨迹；Or the transport vehicle exists on the road, the total number of pedestrians is less than C, the number of minors among the pedestrians is greater than or equal to B but less than D, and the trajectory of at least one of the pedestrians is predicted to invade the trajectory of the transport vehicle;
其中，C、D、B的数量依次减小。Among them, the numbers of C, D, and B decrease in order.
本发明中，ABCD均为经验值，在一个具体实施例中，A为10，B为3，C为20，D为6；即以下情形下判定所述道路安全等级为第一安全等级：In the present invention, ABCD is an empirical value. In a specific embodiment, A is 10, B is 3, C is 20, and D is 6; that is, the road safety level is determined to be the first safety level under the following circumstances:
道路上无所述运输车辆；There are no such transport vehicles on the road;
或者道路上存在所述运输车辆，所述行人的总数量小于10人，且不存在未成年人；Or there are said transport vehicles on the road, the total number of said pedestrians is less than 10, and there are no minors;
或者道路上存在所述运输车辆，所述行人中未成年人的数量小于3，且所述行人的轨迹预测不侵入所述运输车辆轨迹。Or the transportation vehicle exists on the road, the number of minors among the pedestrians is less than 3, and the trajectory prediction of the pedestrians does not invade the transportation vehicle trajectory.
进一步地，以下情形下判定所述道路安全等级为第二安全等级：Further, the road safety level is determined to be the second safety level under the following circumstances:
道路上存在所述运输车辆，所述行人的总数量大于等于10人、小于20人，且不存在未成年人；The transport vehicle exists on the road, the total number of pedestrians is greater than or equal to 10 and less than 20, and there are no minors;
或者道路上存在所述运输车辆，所述行人的总数量小于10人，且所述行人中未成年人的数量大于等于3、小于6；Or the transport vehicle exists on the road, the total number of pedestrians is less than 10, and the number of minors among the pedestrians is greater than or equal to 3 but less than 6;
或者道路上存在所述运输车辆，所述行人中未成年人的数量小于3，且至少一个所述行人的轨迹预测侵入所述运输车辆轨迹。Or the transport vehicle exists on the road, the number of minors among the pedestrians is less than 3, and the trajectory of at least one pedestrian is predicted to invade the trajectory of the transport vehicle.
进一步地，以下情形下判定所述道路安全等级为第三安全等级：Further, the road safety level is determined to be the third safety level under the following circumstances:
道路上存在所述运输车辆，所述行人的总数量大于等于20人，且不存在未成年人；There are said transport vehicles on the road, the total number of said pedestrians is greater than or equal to 20, and there are no minors;
或者道路上存在所述运输车辆，所述行人中未成年人的数量大于等于6，且所述行人的轨迹预测不侵入所述运输车辆轨迹；Or the transport vehicle exists on the road, the number of minors among the pedestrians is greater than or equal to 6, and the trajectory prediction of the pedestrians does not invade the trajectory of the transport vehicle;
或者道路上存在所述运输车辆，所述行人的总数量小于20人，所述行人中未成年人的数量大于等于3、小于6，且至少一个所述行人的轨迹预测侵入所述运输车辆轨迹。Or the transport vehicle exists on the road, the total number of pedestrians is less than 20, the number of minors among the pedestrians is greater than or equal to 3 and less than 6, and the trajectory of at least one pedestrian is predicted to invade the trajectory of the transport vehicle. .
进一步地，当道路安全等级为第一安全等级时，不进行预警提醒；当道路安全等级为第二安全等级时，提醒所述运输车辆减速通过，且提醒所述行人远离道路；当道路安全等级为第三安全等级时，提醒所述运输车辆暂缓通行，提醒所述行人远离道路。Further, when the road safety level is the first safety level, no early warning reminder is provided; when the road safety level is the second safety level, the transport vehicles are reminded to slow down and pass, and the pedestrians are reminded to stay away from the road; when the road safety level is When it is the third safety level, the transport vehicles are reminded to suspend their passage and the pedestrians are reminded to stay away from the road.
本实施例提供的道路安全等级判定具有以下优点：The road safety level determination provided by this embodiment has the following advantages:
危险等级判定自动化：目前已有的物料运输车辆安全监测基本采用传统方式，一些针对车辆夜间出行的少数应用计算机视觉监控的监测也只是当做传统监控手段，辅助人工查看监控。而本发明的区域安全判定方法预先构建了三种区域安全等级，根据道路上的运输车辆以及行人相关信息进行区域安全的自动化确定；Automation of hazard level determination: Currently, the existing safety monitoring of material transport vehicles basically adopts traditional methods. Some of the few computer vision monitoring applications for vehicles traveling at night are only used as traditional monitoring methods to assist manual viewing and monitoring. The regional safety determination method of the present invention pre-constructs three regional safety levels, and automatically determines regional safety based on relevant information about transportation vehicles and pedestrians on the road;
信息利用全面：道路安全等级中囊括了多类信息，包括车辆、行人数量信息；人员组成信息及轨迹预测信息等，相比传统的人工观察监控法具有信息获取全面、状况判断准确、决策建议及时的优点。道路安全等级的设置通过预设的标准，结合自动化的识别技术实现了检测、分析、预警全流程的智能化、标准化，真正提高了物料运输车穿越城镇的安全性，进一步地保障了工程的顺利进行。Comprehensive information utilization: The road safety level includes many types of information, including vehicle and pedestrian number information; personnel composition information and trajectory prediction information, etc. Compared with the traditional manual observation and monitoring method, it has comprehensive information acquisition, accurate situation judgment, and timely decision-making suggestions. The advantages. The setting of road safety levels is based on preset standards and combined with automated identification technology to achieve intelligentization and standardization of the entire process of detection, analysis, and early warning, which truly improves the safety of material transport vehicles traveling through towns and further ensures the smoothness of the project. conduct.
在另一实施例中，本发明还提供一种运输车辆穿越道路时的交通预警系统，该交通预警系统用于实现上述任一项所述的交通预警方法，可与上述交通预警方法对应参照理解，该交通预警系统包括：In another embodiment, the present invention also provides a traffic warning system for transport vehicles crossing roads. The traffic warning system is used to implement any of the above-mentioned traffic warning methods, and can be understood by reference to the above-mentioned traffic warning methods. , the traffic warning system includes:
监测模块，用于对道路状态进行监测，获取道路监控信息；Monitoring module, used to monitor road status and obtain road monitoring information;
分析处理模块，用于根据所述道路监控信息，获取道路上的运输车辆以及行人相关信息，其中道路上的运输车辆以及行人相关信息包括所述运输车辆的数量和位置信息、所述行人的数量和位置信息、所述行人的人员组成信息以及所述运输车辆和所述行人的轨迹入侵预测信息；An analysis and processing module configured to obtain information related to transportation vehicles and pedestrians on the road based on the road monitoring information, where the information related to transportation vehicles and pedestrians on the road includes the number and location information of the transportation vehicles and the number of pedestrians. and location information, the pedestrian composition information, and the trajectory intrusion prediction information of the transportation vehicle and the pedestrian;
安全判定模块，用于根据道路上的运输车辆以及行人相关信息判断道路安全等级，其中所述道路安全等级预先设置为包括第一安全等级、第二安全等级和第三安全等级三个级别；A safety determination module, used to determine the road safety level based on information related to transport vehicles and pedestrians on the road, where the road safety level is preset to include three levels: a first safety level, a second safety level and a third safety level;
预警模块，用于根据所述道路安全等级对所述运输车辆和/或所述行人进行预警提醒。An early warning module is used to provide early warning to the transportation vehicle and/or the pedestrian according to the road safety level.
具体地，监测模块可为摄像头等监控设备，可安装在道路上，以能对道路状态进行全面监控为目的，具体位置和数量不做限定。监测模块用于实时监测物料运输车辆通过道路的全过程；分析处理模块借助图像处理、目标检测和跟踪等算法，可以实现对车辆和行人的实时监测和轨迹预测。Specifically, the monitoring module can be a monitoring device such as a camera, which can be installed on the road for the purpose of comprehensively monitoring the road status. The specific location and quantity are not limited. The monitoring module is used to monitor the entire process of material transport vehicles passing through the road in real time; the analysis and processing module uses algorithms such as image processing, target detection and tracking to achieve real-time monitoring and trajectory prediction of vehicles and pedestrians.
进一步地，所述分析处理模块具体包括：Further, the analysis and processing module specifically includes:
目标检测模块，用于根据所述道路监控信息，获取所述运输车辆的数量和位置信息、所述行人的数量和位置信息以及所述行人的人员组成信息；A target detection module, configured to obtain the number and location information of the transport vehicles, the number and location information of the pedestrians, and the personnel composition information of the pedestrians based on the road monitoring information;
时序分析模块，用于根据所述道路监控信息，获取所述运输车辆和所述行人的轨迹入侵预测信息。A time series analysis module, configured to obtain trajectory intrusion prediction information of the transportation vehicle and the pedestrian based on the road monitoring information.
分析处理模块内置目标检测及轨迹预测等多类计算机视觉算法，用于确定车辆、行人位置信息，出行居民人员组成信息及行人轨迹进行时空预测。The analysis and processing module has built-in computer vision algorithms such as target detection and trajectory prediction, which are used to determine the location information of vehicles and pedestrians, the composition information of traveling residents, and pedestrian trajectories for spatio-temporal prediction.
在一个具体实施例中，本发明提供一种基于计算机视觉技术的施工物料运输车辆穿越乡镇道路的交通安全监测提醒预警系统及方法，以实现对小镇区域的安全监控以及及时提醒，减少施工物料运输车辆穿越城镇发生事故的可能性，提高物料运输车辆及小镇居民的出行安全。In a specific embodiment, the present invention provides a traffic safety monitoring reminder and early warning system and method for construction material transport vehicles crossing township roads based on computer vision technology to achieve safety monitoring and timely reminders in small town areas and reduce construction materials The possibility of accidents involving transport vehicles crossing towns improves the travel safety of material transport vehicles and town residents.
参考图2，该预警系统包括车辆及行人状态监测系统、区域安全判定系统和安全状况预警系统，其中车辆及行人状态监测系统用于识别车辆、行人的特征信息，包括位置信息、未成年数量信息、人员聚集信息等；并对行人运动状态信息做短时预测；区域安全判定系统接收车辆及行人状态监测系统分析出的区域信息，根据上述系统可获取的信息类别建立区域安全判定方法，据此方法判定车辆穿越城镇时的安全状况，并给出实时区域安全等级与安全建议；安全状况预警系统根据区域安全判定系统给出的实时区域安全等级与安全建议同步提醒车辆行驶状态及广播小镇居民安全出行。Referring to Figure 2, the early warning system includes a vehicle and pedestrian status monitoring system, a regional safety determination system and a safety status early warning system. The vehicle and pedestrian status monitoring system is used to identify the characteristic information of vehicles and pedestrians, including location information and number of minors. , people gathering information, etc.; and make short-term predictions of pedestrian movement status information; the regional safety determination system receives the regional information analyzed by the vehicle and pedestrian status monitoring system, and establishes a regional safety determination method based on the information categories that can be obtained by the above system. The method determines the safety status of vehicles when passing through towns and provides real-time regional safety levels and safety suggestions; the safety status early warning system simultaneously reminds vehicle driving status and broadcasts to town residents based on the real-time regional safety levels and safety suggestions given by the regional safety determination system. Travel safely.
车辆及行人状态监测系统包括监测模块和分析处理模块，监测模块用于实时视频监控小镇车辆、人员信息，并将记录车辆、人员信息的视频图像发送至分析处理模块。分析处理模块用于接收视频图像信息，并通过目标检测模块将记录的视频图像信息生成数字信息，具体包括车辆、行人位置信息，区域人员组成信息(成年人/未成年人数量)，人员数量信息；并通过时序分析模块对车辆及行人运动轨迹进行短时预测。The vehicle and pedestrian status monitoring system includes a monitoring module and an analysis and processing module. The monitoring module is used for real-time video monitoring of vehicle and personnel information in the town, and sends video images recording vehicle and personnel information to the analysis and processing module. The analysis and processing module is used to receive video image information, and generate digital information from the recorded video image information through the target detection module, specifically including vehicle and pedestrian location information, regional personnel composition information (number of adults/minors), and number of personnel information. ; And use the time series analysis module to make short-term predictions of vehicle and pedestrian movement trajectories.
分析处理模块包括目标检测模块及时序分析模块，分析处理模块基于目标检测技术识别出车辆及行人的位置信息，进而判断出人员的聚集状态；同时检测未成年人数量；并基于时序分析预测方法transformer对车辆及行人轨迹进行短时预测。The analysis and processing module includes a target detection module and a sequence analysis module. The analysis and processing module identifies the location information of vehicles and pedestrians based on target detection technology, and then determines the gathering status of people; simultaneously detects the number of minors; and based on the sequence analysis prediction method transformer Short-term prediction of vehicle and pedestrian trajectories.
在一个具体实施例中，目标检测模块需要在有较高检测精度的基础上实现实时检测的能力，本发明对比了以精度为代表的faster R-CNN算法和以速度为代表的Yolo系列算法，提出了以Transformer作为特征提取网络，Yolo作为目标检测网络的Transformer+Yolo目标检测算法。In a specific embodiment, the target detection module needs to have the ability to achieve real-time detection on the basis of higher detection accuracy. The present invention compares the faster R-CNN algorithm represented by accuracy and the Yolo series algorithm represented by speed. A Transformer+Yolo target detection algorithm using Transformer as the feature extraction network and Yolo as the target detection network is proposed.
时序分析模块中，为了克服传统RNN难以对较长序列进行建模的问题，同时提升时序建模能力，本发明采用基于Transformer的时间-空间图卷积(Spatio-Temporal GraphTransformer)模型解决轨迹预测问题。有研究表明该方法在解决轨迹预测问题上表现良好。为了进行行人轨迹预测，首先，沿用传统Transformer对行人轨迹进行时序建模；其次，设计基于Transformer的通用图卷积算法，TGConv；最后，交替使用时序以及空间Transformer完成对时间-空间图的建模。其中，对于时序建模，使用一个TransformerEncoder：In the timing analysis module, in order to overcome the difficulty of traditional RNN in modeling longer sequences and improve timing modeling capabilities, the present invention uses a Transformer-based Spatio-Temporal GraphTransformer model to solve the trajectory prediction problem. . Some studies have shown that this method performs well in solving trajectory prediction problems. In order to predict pedestrian trajectories, firstly, the traditional Transformer is used to model pedestrian trajectories in time series; secondly, a general graph convolution algorithm based on Transformer, TGConv, is designed; finally, time series and spatial Transformers are used alternately to complete the modeling of time-space graph. . Among them, for timing modeling, a TransformerEncoder is used:
whereheadj＝Attj(Qi,Ki,Vi)；whereheadj =Attj (Qi ,Ki ,Vi );
Att(Qi,Ki,Vi)是自注意力，反应了当前位置与其他所有位置的关系强度。其中，Qi,Ki,Vi分别代表是query，key和value矩阵；dk是key的维度；Softmax(QiKiT)计算了不同时刻之间的注意力权重。Att(Qi ,Ki ,Vi ) is self-attention, which reflects the strength of the relationship between the current position and all other positions. Among them, Qi , Ki , Vi represent query, key and value matrices respectively; dk is the dimension of key; Softmax (Qi KiT ) calculates the attention weight between different moments.
MultiHead(Qi,Ki,Vi)是多头注意力。其中，表示k个不同的注意力头；k可为行人总数量；fo是输出函数，是一个全连接层；headj是第j个注意力头的输出。MultiHead(Qi ,Ki ,Vi ) is multi-head attention. in, Represents k different attention heads; k can be the total number of pedestrians; fo is the output function, which is a fully connected layer; headj is the output of the jth attention head.
对于第i号行人，将其历史位置信息组合成T×2的矩阵(T为时间长度，2代表x和y坐标)，通过三个不同的全链接网络提取出不同的特征矩阵。通过上式，可以计算对于行人i的不同时刻的位置之间的自注意力，以此提取出行人i的时间维度运动模式，并且完成时序建模。For the i-th pedestrian, its historical location information is combined into a T×2 matrix (T is the time length, 2 represents the x and y coordinates), and different feature matrices are extracted through three different full-link networks. Through the above formula, the self-attention between the positions of pedestrian i at different moments can be calculated, thereby extracting the time dimension motion pattern of pedestrian i and completing time series modeling.
对于空间建模，对于每一时刻t，根据行人位置信息构建一个空间图：若两个行人的空间距离小于d，即认为这两个行人之间存在边。为了对行人复杂的交互进行较好的注意力机制建模，采用Transformer Encoder作为基础算子。自注意力机制实际上可以被理解为全链接图上的图卷积。具体而言，对于任意一个特征向量hi，提取其query向量qi，key向量ki，以及value向量vi。那么可以定义图卷积中从j节点到i节点的message passing为：For spatial modeling, for each time t, a spatial graph is constructed based on the pedestrian position information: if the spatial distance between two pedestrians is less than d, it is considered that there is an edge between the two pedestrians. In order to better model the attention mechanism for complex pedestrian interactions, Transformer Encoder is used as the basic operator. The self-attention mechanism can actually be understood as graph convolution on a fully linked graph. Specifically, for any feature vector hi , extract its query vector qi , key vector ki , and value vector vi . Then the message passing from j node to i node in graph convolution can be defined as:
mj→i表示了节点i的query与节点j的key之间的点积。这个点积捕获了节点i与节点j之间的关系强度。以此为基础，将Transformer的自注意力机制用message passing重新定义：mj→i represents the dot product between the query of node i and the key of node j. This dot product captures the strength of the relationship between node i and node j. Based on this, Transformer's self-attention mechanism is redefined using message passing:
是每个节点的value矩阵的列表，与Softmax后的权重相乘，可以得到加权的value；n可为行人总数量。基于上述观察，得出TGConv(Transformer-based GraphConvolution)： is a list of value matrices for each node, multiplied by the weight after Softmax, the weighted value can be obtained; n can be the total number of pedestrians. Based on the above observations, TGConv (Transformer-based GraphConvolution) is derived:
hi′＝fout(Att(i))+Att(i)；hi ′=fout (Att(i))+Att(i);
其中，Nb(i)代表i节点的邻接节点；hi是节点i(在此代表行人)的特征向量；hi′是行人i的历史轨迹信息、与其他行人的相对位置和交互信息的融合表示。TGConv将Transformer中的注意力机制应用到了任意图卷积，更好地解决了复杂行人交互的空间建模问题。Among them, Nb(i) represents the adjacent nodes of node i; hi is the feature vector of node i (representing the pedestrian here); hi ′ is the fusion of historical trajectory information of pedestrian i, relative position and interaction information with other pedestrians express. TGConv applies the attention mechanism in Transformer to arbitrary graph convolution to better solve the spatial modeling problem of complex pedestrian interaction.
本方法使用两个结构不同的encoder进行信息提取和建模。在第一个encoder中，历史信息分别被一个时间和一个空间Transformer进行处理，提取特征并进行融合。在第二个encoder中，先进行空间信息提取，将上一步提取过的时间和空间信息进行融合，再用一个时序Transformer进行最后的预测。This method uses two encoders with different structures for information extraction and modeling. In the first encoder, historical information is processed by a temporal and spatial Transformer respectively, features are extracted and fused. In the second encoder, spatial information is extracted first, the time and space information extracted in the previous step are fused, and then a time series Transformer is used for the final prediction.
区域安全判定系统包括安全判定模块，接收车辆及行人状态监测系统发出的车辆、行人的特征信息及行人运动状态信息建立区域安全判定方法，据此方法判定车辆穿越小镇时的安全状况，并给出实时区域安全等级与安全建议；区域安全判定方法基于分析处理模块得出的车辆及人员信息构建；所述区域安全分为三个等级，其中一级区域安全即第一安全等级表示安全程度高，对行人可不做提醒；二级区域安全即第二安全等级表示安全程度一般，物料运输车辆需减速观察后通过；三级区域安全即第三安全等级表示安全程度较低，物料运输车辆需暂缓通过。三个区域安全等级确定需要根据车辆及行人状态监测系统的分析处理模块得出的数字信息，结合所述的区域安全判定系统中的三级区域安全等级判定方法判断物料运输车辆穿越小镇时的区域安全性。The regional safety determination system includes a safety determination module, which receives vehicle and pedestrian characteristic information and pedestrian movement status information from the vehicle and pedestrian status monitoring system to establish a regional safety determination method. Based on this method, the safety status of vehicles passing through a small town is determined, and the Produce real-time regional safety levels and safety suggestions; the regional safety determination method is constructed based on the vehicle and personnel information obtained by the analysis and processing module; the regional security is divided into three levels, among which the first-level regional security, that is, the first security level, indicates a high degree of security , there is no need to remind pedestrians; Level 2 regional safety, or the second safety level, means the level of safety is average, and material transport vehicles need to slow down and observe before passing; Level 3 area safety, or the third safety level, means the level of safety is low, and material transport vehicles need to be delayed. pass. The determination of the three regional safety levels needs to be based on the digital information obtained by the analysis and processing module of the vehicle and pedestrian status monitoring system, combined with the three-level regional safety level determination method in the regional safety determination system to determine the safety of material transport vehicles when crossing the town. Regional security.
安全状况预警系统包括预警模块，具体可包括物料运输车辆语音提醒模块、城镇广播提醒模块，其中，物料运输车辆语音提醒模块接收区域安全等级并语音播报行驶建议；小镇广播提醒模块同样接收区域安全等级及小镇居民出行建议，并语音播报提醒小镇居民安全出行。安全状况预警系统根据上述判定结果对车辆驾驶员及小镇居民进行安全状况语音播报预警，同时给出车辆驾驶建议。物料运输车辆语音提醒模块、城镇广播提醒模块可设置在道路上。The safety status early warning system includes an early warning module, which can specifically include a material transportation vehicle voice reminder module and a town broadcast reminder module. Among them, the material transportation vehicle voice reminder module receives the regional safety level and broadcasts driving suggestions by voice; the town radio reminder module also receives regional safety Levels and travel recommendations for town residents, and voice broadcasts to remind town residents to travel safely. Based on the above determination results, the safety status early warning system provides voice broadcast warnings of safety status to vehicle drivers and town residents, and also provides vehicle driving suggestions. The material transport vehicle voice reminder module and the town broadcast reminder module can be set up on the road.
在一个具体实施例中，当城镇区域安全状态为一级安全区域时，安全状况预警系统不做任何动作，其余各系统常规运行；当城镇区域安全状态为二级安全区域时，物料运输车辆语音提醒模块播报：“您正在穿越城镇，人员较多，请减速观察，缓慢通过！”，小镇广播提醒模块播报：“施工物料运输车辆正在穿越城镇，请远离道路中心，看管好您的孩子！”；当城镇区域安全状态为三级安全区域时，物料运输车辆语音提醒模块播报：“您正在穿越城镇，人员密集，请暂缓通行！”，小镇广播提醒模块播报：“施工物料运输车辆正在穿越城镇，请勿聚集，请各位家长带走自己的孩子，请勿逗留！”In a specific embodiment, when the safety status of the urban area is a first-level safety area, the safety status early warning system does not take any action, and the other systems operate normally; when the safety status of the urban area is a second-level safety area, the material transport vehicle voice The reminder module broadcast: "You are passing through a town. There are many people. Please slow down and observe and pass slowly!". The town broadcast reminder module broadcast: "Construction material transport vehicles are passing through the town. Please stay away from the center of the road and watch your children!" "; When the security status of the urban area is a third-level safety area, the material transport vehicle voice reminder module broadcasts: "You are passing through the town, and there are dense people, please suspend the passage!", the town radio reminder module broadcast: "Construction material transport vehicles are passing through. Traveling across town, please do not gather, parents please take your children and please do not linger!”
一种基于计算机视觉技术的施工物料运输车辆穿越城镇的交通安全监测提醒预警方法，具体步骤包括：A traffic safety monitoring and early warning method for construction material transport vehicles crossing towns based on computer vision technology. The specific steps include:
监测乡镇区域道路上的施工物料运输车辆及行人状态，识别车辆、行人的位置信息，人员组成信息(未成年人数)，并且预测行人短时轨迹；Monitor the status of construction material transport vehicles and pedestrians on roads in townships and towns, identify the location information of vehicles and pedestrians, personnel composition information (number of minors), and predict short-term trajectories of pedestrians;
建立区域安全等级判定方法，结合车辆及行人状态，判断区域安全等级；并生成不同安全等级下的车辆及居民出行建议；Establish a method to determine the regional safety level, combined with the status of vehicles and pedestrians, to determine the regional safety level; and generate travel suggestions for vehicles and residents under different safety levels;
安全状况预警，根据区域安全等级分别通过物料运输车辆语音提醒模块、小镇广播提醒模块实时播报小镇居民出行建议，给出相关安全提醒。Safety status early warning: according to the regional safety level, the material transportation vehicle voice reminder module and the town broadcast reminder module broadcast real-time travel suggestions for small town residents and give relevant safety reminders.
本发明从两个方面助力管控施工物料运输车辆运输过程的安全性。第一个方面从驾驶人员角度出发，将监控摄像头置于物料运输车辆穿越乡镇的路径之上，采集到车辆及小镇居民出行信息，如车辆居民位置、人员组成，再通过分析处理模块的算法进行短时轨迹预测，根据分析处理模块计算结果结合提前制定的区域安全判定方法确定区域安全等级，并通过物料运输车辆语音提醒模块给出司机通过城镇区域的驾驶建议，如“您正在穿越城镇，人员密集，请暂缓通行！”，直到车辆驶离小镇区域。The invention helps manage and control the safety of the transportation process of construction material transportation vehicles from two aspects. The first aspect is from the perspective of drivers. Surveillance cameras are placed on the paths of material transport vehicles passing through towns and villages, and travel information of vehicles and town residents, such as the location of vehicle residents and personnel composition, is collected, and then through the algorithm of the analysis and processing module Carry out short-term trajectory prediction, determine the regional safety level based on the calculation results of the analysis and processing module combined with the regional safety determination method formulated in advance, and give the driver driving suggestions through urban areas through the material transportation vehicle voice reminder module, such as "You are crossing the town. It’s crowded, please hold off!” until the vehicle leaves the town area.
第二个方面从小镇出行居民角度出发，在确定了区域安全等级后，通过小镇广播提醒模块给出小镇居民出行建议，如“施工物料运输车辆正在穿越城镇，请勿聚集，请各位家长带走自己的孩子，请勿逗留！”，直到车辆驶离小镇区域。The second aspect is from the perspective of residents traveling in small towns. After determining the regional safety level, travel suggestions for small town residents are given through the town broadcast reminder module, such as "Construction material transport vehicles are passing through the town. Please do not gather. Please Parents, take your children away and don’t stay!” until the vehicle leaves the town area.
本发明提供的一种施工物料运输车辆穿越乡镇道路的实时交通安全监测提醒预警方法及系统，通过车辆及行人状态监测系统非接触式检测车辆及行人的数量、位置信息，同时对行人运动轨迹做实时预测，对区域安全等级的判定具有全局性、实时性和预测性；区域安全判定系统预设三级安全区域，并明确出每一类的实用条件，实现了区域安全判定的自动化；安全状况预警系统将根据前述系统的判定结果，实时对运料运输车辆及小镇居民给出安全预警。与现有的监控手段相比，该系统及方法具备自动化的信息采集及分析能力，量化了区域交通安全实时等级，提高了施工物料运输车辆穿越城镇的安全性。The invention provides a real-time traffic safety monitoring reminder and early warning method and system for construction material transport vehicles crossing township roads. The number and location information of vehicles and pedestrians are non-contactly detected through the vehicle and pedestrian status monitoring system, and the movement trajectories of pedestrians are measured at the same time. Real-time prediction, the judgment of regional safety level is global, real-time and predictive; the regional safety judgment system presets three-level safety areas and clarifies the practical conditions of each category, realizing the automation of regional safety judgment; safety status The early warning system will give real-time safety warnings to material transport vehicles and town residents based on the judgment results of the aforementioned system. Compared with existing monitoring methods, this system and method has automated information collection and analysis capabilities, quantifies the real-time level of regional traffic safety, and improves the safety of construction material transport vehicles crossing towns.
进一步地，本发明基于计算机视觉技术，利用普通监控摄像头拍摄物料运输车辆穿越城镇区域的视频，并根据基于Yolo v8的目标检测算法及基于Transformer的时空轨迹预测算法获取车辆及行人的特征信息及预测信息，不同于传统监控技术只能用来获取监控视频，区域安全自动化判别及预警方式具备一定的交通事件理解能力，可根据预设算法做出自动化预警，大大提高了物料运输车辆安全通过乡镇的可靠性。Furthermore, the present invention is based on computer vision technology, using ordinary surveillance cameras to capture videos of material transport vehicles crossing urban areas, and obtaining characteristic information and predictions of vehicles and pedestrians based on the Yolo v8-based target detection algorithm and the Transformer-based spatio-temporal trajectory prediction algorithm. Information, unlike traditional surveillance technology which can only be used to obtain surveillance videos, the regional safety automated identification and early warning method has a certain ability to understand traffic events and can make automated early warnings based on preset algorithms, which greatly improves the safety of material transport vehicles passing through towns and villages. reliability.
本领域的技术人员容易理解，以上所述仅为本发明的较佳实施例而已，并不用以限制本发明，凡在本发明的精神和原则之内所作的任何修改、等同替换和改进等，均应包含在本发明的保护范围之内。It is easy for those skilled in the art to understand that the above descriptions are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions and improvements, etc., made within the spirit and principles of the present invention, All should be included in the protection scope of the present invention.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311321519.5ACN117292578A (en) | 2023-10-12 | 2023-10-12 | Traffic early warning method and system for transport vehicle crossing road |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311321519.5ACN117292578A (en) | 2023-10-12 | 2023-10-12 | Traffic early warning method and system for transport vehicle crossing road |
| Publication Number | Publication Date |
|---|---|
| CN117292578Atrue CN117292578A (en) | 2023-12-26 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311321519.5APendingCN117292578A (en) | 2023-10-12 | 2023-10-12 | Traffic early warning method and system for transport vehicle crossing road |
| Country | Link |
|---|---|
| CN (1) | CN117292578A (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118015701A (en)* | 2024-02-01 | 2024-05-10 | 云南大学 | A passenger behavior recognition method for subway platforms with high-density passenger flow scenarios |
| CN119251750A (en)* | 2024-08-14 | 2025-01-03 | 南京聚聚网络科技有限公司 | An intelligent urban dog management platform based on the Internet of Things |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118015701A (en)* | 2024-02-01 | 2024-05-10 | 云南大学 | A passenger behavior recognition method for subway platforms with high-density passenger flow scenarios |
| CN119251750A (en)* | 2024-08-14 | 2025-01-03 | 南京聚聚网络科技有限公司 | An intelligent urban dog management platform based on the Internet of Things |
| CN119251750B (en)* | 2024-08-14 | 2025-08-22 | 江苏聚聚网络科技有限公司 | An urban intelligent dog management platform based on the Internet of Things |
| Publication | Publication Date | Title |
|---|---|---|
| KR102122859B1 (en) | Method for tracking multi target in traffic image-monitoring-system | |
| CN112633057B (en) | Intelligent monitoring method for abnormal behavior in bus | |
| CN117292578A (en) | Traffic early warning method and system for transport vehicle crossing road | |
| KR102122850B1 (en) | Solution for analysis road and recognition vehicle license plate employing deep-learning | |
| CN111462488A (en) | An Intersection Safety Risk Assessment Method Based on Deep Convolutional Neural Network and Intersection Behavior Feature Model | |
| CN112686090A (en) | Intelligent monitoring system for abnormal behaviors in bus | |
| KR102174556B1 (en) | Apparatus for monitoring image to control traffic information employing Artificial Intelligence and vehicle number | |
| CN101266710A (en) | A rule-based all-weather intelligent video analysis and monitoring method | |
| KR102403170B1 (en) | Apparatus for monitoring image employing to detect of vehicle number and cotrolling device | |
| CN103208190B (en) | Traffic flow detection method based on object detection | |
| CN113076821A (en) | Event detection method and device | |
| CN117975713A (en) | Intelligent traffic guiding management system for industrial upstairs | |
| CN113870551B (en) | A roadside monitoring system capable of distinguishing between dangerous and non-dangerous driving behavior | |
| WO2018209470A1 (en) | License plate identification method and system | |
| Hnoohom et al. | The video-based safety methodology for pedestrian crosswalk safety measured: The case of Thammasat University, Thailand | |
| Wang et al. | Vision-based highway traffic accident detection | |
| Phatangare et al. | Real-time traffic management using deep learning and object detection using YOLOv8 | |
| CN113283309A (en) | Method and system for identifying road construction state according to real-time pavement image | |
| CN209486891U (en) | Shared bicycle illegal parking detection system | |
| Girija et al. | Road traffic accident prediction using deep learning | |
| Shankar et al. | Automated Traffic Management System Using Deep Learning Based Object Detection | |
| Zhang et al. | Machine learning and computer vision-enabled traffic sensing data analysis and quality enhancement | |
| Yoo et al. | Truck detection and counting in low-light condition: do we need infrared camera? | |
| Upadhyay et al. | Traffic Monitoring System using YOLOv3 Model | |
| Patil et al. | Enhancing Traffic Management with Deep Learning Based Vehicle Detection and Scheduling Systems |
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |