CN114783172A - A method and system for identifying an empty parking space in a parking lot, and a computer-readable storage medium - Google Patents
A method and system for identifying an empty parking space in a parking lot, and a computer-readable storage mediumDownload PDFInfo
- Publication number
- CN114783172A CN114783172ACN202110088091.9ACN202110088091ACN114783172ACN 114783172 ACN114783172 ACN 114783172ACN 202110088091 ACN202110088091 ACN 202110088091ACN 114783172 ACN114783172 ACN 114783172A
- Authority
- CN
- China
- Prior art keywords
- parking space
- point cloud
- empty
- point
- simulated
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/14—Traffic control systems for road vehicles indicating individual free spaces in parking areas
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Traffic Control Systems (AREA)
Abstract
Description
Translated fromChinese技术领域technical field
本发明涉及自动泊车技术领域，具体涉及一种停车场空车位识别方法及系统、计算机可读存储介质。The present invention relates to the technical field of automatic parking, in particular to a method and system for identifying an empty parking space in a parking lot, and a computer-readable storage medium.
背景技术Background technique
空车位定位是自动泊车技术中的基础；目前车位识别方法主要可分为基于超声波雷达、基于地磁传感器、基于环视图像等三种车位识别方法；其中，基于超声波雷达的车位识别方法是利用超声波雷达对周围环境(车辆、障碍物等)进行感知来识别车位，但超声波雷达信息点数较少，无法对车位进行准确识别，只能大概推断可行驶或无障碍区域；基于地磁传感器的车位识别方法需要提前实现对停车区域的整体改造，其应用不方便；基于环视图像的车位识别方法是通过环视摄像头进行实时图像采集，对采集到的图像中的车位进行提取，并结合对周围环境的感知结果确定目标空车位，但其依赖于车位线的识别，如果车位的车位线不明显或不完整，则识别效果较差。Empty parking space positioning is the basis of automatic parking technology; the current parking space recognition methods can be mainly divided into three types of parking space recognition methods based on ultrasonic radar, based on geomagnetic sensors, and based on surround view images; among them, the parking space recognition method based on ultrasonic radar uses ultrasonic waves. Radar perceives the surrounding environment (vehicles, obstacles, etc.) to identify the parking space, but the ultrasonic radar has few information points, so it cannot accurately identify the parking space, and can only roughly infer the drivable or barrier-free area; the parking space recognition method based on the geomagnetic sensor It is necessary to realize the overall transformation of the parking area in advance, and its application is inconvenient; the parking space recognition method based on the surround view image is to collect the real-time image through the surround view camera, extract the parking space in the collected image, and combine the perception results of the surrounding environment. Determine the target empty parking space, but it depends on the recognition of the parking space line. If the parking space line of the parking space is not obvious or incomplete, the recognition effect is poor.
综上，有必要提出一种新的车位识别方法，以克服现有基于超声波雷达、基于地磁传感器、基于环视图像等三种车位识别方法所存在的上述技术缺陷。To sum up, it is necessary to propose a new parking space recognition method to overcome the above-mentioned technical shortcomings of the existing three parking space recognition methods based on ultrasonic radar, based on geomagnetic sensors, and based on surround view images.
发明内容SUMMARY OF THE INVENTION
本发明的目的在于提出一种停车场空车位识别方法及系统、计算机可读存储介质，以克服现有基于超声波雷达、基于地磁传感器、基于环视图像等三种车位识别方法所存在的上述技术缺陷。The purpose of the present invention is to propose a parking space identification method and system, and a computer-readable storage medium to overcome the above-mentioned technical defects of the existing three parking space identification methods based on ultrasonic radar, geomagnetic sensor, and surround view image. .
为实现上述目的，本发明第一方面提出一种停车场空车位识别方法，包括：In order to achieve the above object, a first aspect of the present invention proposes a method for identifying an empty parking space in a parking lot, comprising:
获取车辆行驶过程中车载深度相机对车辆单侧方向进行探测获得的深度图像；Obtain the depth image obtained by the vehicle-mounted depth camera detecting the unilateral direction of the vehicle during the driving process of the vehicle;
对所述深度图像进行点云转换获得第一点云；Perform point cloud conversion on the depth image to obtain a first point cloud;
按照预设滤波规则对所述第一点云进行滤波获得第二点云；Filter the first point cloud according to a preset filtering rule to obtain a second point cloud;
对所述第二点云进行下采样获得第三点云；down-sampling the second point cloud to obtain a third point cloud;
根据所述第三点云以及预设的车位三维空间参数识别所述第三点云中是否存在空车位；若存在空车位，则生成空车位的车位框。Identify whether there is an empty parking space in the third point cloud according to the third point cloud and the preset three-dimensional space parameters of the parking space; if there is an empty parking space, a parking space frame of the empty parking space is generated.
可选地，所述按照预设滤波规则对所述第一点云进行滤波获得第二点云，包括：Optionally, filtering the first point cloud to obtain a second point cloud according to a preset filtering rule includes:
进行一次滤波，以过滤掉所述第一点云中深度值等于预设深度值的点；Perform a filtering to filter out the points whose depth value is equal to the preset depth value in the first point cloud;
进行二次滤波，以过滤掉所述第一点云中的离群点。Secondary filtering is performed to filter out outliers in the first point cloud.
可选地，所述对所述第二点云进行下采样获得第三点云，包括：Optionally, performing downsampling on the second point cloud to obtain a third point cloud includes:
把所述第二点云划分成多个立方体，每个立方体中保留1个点，从而获得第三点云。The second point cloud is divided into a plurality of cubes, and one point is reserved in each cube, thereby obtaining a third point cloud.
可选地，所述根据所述第三点云以及预设的车位三维空间参数识别所述深度图像中是否存在空车位，包括：Optionally, identifying whether there is an empty parking space in the depth image according to the third point cloud and a preset three-dimensional space parameter of a parking space includes:
遍历所述第三点云的点云空间中所有点，将每一点作为车位框左上角点，并根据该车位框左上角点以及所述预设的车位三维空间参数生成每一点所对应的模拟车位框，判断该模拟车位框中是否存在至少一个所述第三点云的点；若是，则该模拟车位框为无效模拟车位框；若否，则该模拟车位框为有效模拟车位框；Traverse all points in the point cloud space of the third point cloud, take each point as the upper left corner of the parking space frame, and generate a simulation corresponding to each point according to the upper left corner point of the parking space frame and the preset three-dimensional space parameters of the parking space a parking space frame, determine whether there is at least one point of the third point cloud in the simulated parking space frame; if so, the simulated parking space frame is an invalid simulated parking space frame; if not, the simulated parking space frame is a valid simulated parking space frame;
统计所有有效模拟车位框所对应的点，并根据所述所有有效模拟车位框所对应的点生成空车位的车位框。Points corresponding to all valid simulated parking space frames are counted, and a parking space frame of an empty parking space is generated according to the points corresponding to all valid simulated parking space frames.
可选地，所述方法还包括：Optionally, the method further includes:
根据所述第一点云、第二点云、第三点云以及所述车位框分别生成第一显示信息、第二显示信息、第三显示信息、第四显示信息，并将所述第一显示信息、第二显示信息、第三显示信息、第四显示信息发送至车载显示单元进行同步显示。According to the first point cloud, the second point cloud, the third point cloud and the parking space frame, first display information, second display information, third display information, and fourth display information are respectively generated, and the first display information is The display information, the second display information, the third display information, and the fourth display information are sent to the vehicle-mounted display unit for synchronous display.
本发明第二方面提出一种停车场空车位识别系统，包括：A second aspect of the present invention proposes an empty parking space identification system in a parking lot, comprising:
图像获取单元，用于获取车辆行驶过程中车载深度相机对车辆单侧方向进行探测获得的深度图像；The image acquisition unit is used to acquire the depth image obtained by the vehicle-mounted depth camera detecting the unilateral direction of the vehicle during the driving of the vehicle;
点云转换单元，用于对所述深度图像进行点云转换获得第一点云；a point cloud conversion unit, configured to perform point cloud conversion on the depth image to obtain a first point cloud;
点云过滤单元，用于按照预设滤波规则对所述第一点云进行滤波获得第二点云；a point cloud filtering unit, configured to filter the first point cloud according to a preset filtering rule to obtain a second point cloud;
点云采样单元，用于对所述第二点云进行下采样获得第三点云；以及a point cloud sampling unit, configured to downsample the second point cloud to obtain a third point cloud; and
空车位识别单元，用于根据所述第三点云以及预设的车位三维空间参数识别所述第三点云中是否存在空车位；若存在空车位，则生成空车位的车位框。The empty parking space identification unit is used for identifying whether there is an empty parking space in the third point cloud according to the third point cloud and the preset three-dimensional space parameters of the parking space; if there is an empty parking space, a parking space frame of the empty parking space is generated.
可选地，所述点云过滤单元，包括：Optionally, the point cloud filtering unit includes:
第一滤波单元，用于进行一次滤波，以过滤掉所述第一点云中深度值等于预设深度值的点；a first filtering unit, configured to perform a filtering to filter out points whose depth value is equal to a preset depth value in the first point cloud;
第二滤波单元，用于进行二次滤波，以过滤掉所述第一点云中的离群点。The second filtering unit is configured to perform secondary filtering to filter out outliers in the first point cloud.
可选地，所述点云采样单元，具体用于Optionally, the point cloud sampling unit is specifically used for
把所述第二点云划分成多个立方体，每个立方体中保留1个点，从而获得第三点云。The second point cloud is divided into a plurality of cubes, and one point is reserved in each cube, thereby obtaining a third point cloud.
可选地，所述空车位识别单元，具体包括：Optionally, the empty parking space identification unit specifically includes:
模拟单元，用于遍历所述第三点云的点云空间中所有点，将每一点作为车位框左上角点，并根据该车位框左上角点以及所述预设的车位三维空间参数生成每一点所对应的模拟车位框，判断该模拟车位框中是否存在至少一个所述第三点云的点；若是，则该模拟车位框为无效模拟车位框；若否，则该模拟车位框为有效模拟车位框；The simulation unit is used to traverse all the points in the point cloud space of the third point cloud, take each point as the upper left corner point of the parking space frame, and generate each point according to the upper left corner point of the parking space frame and the preset three-dimensional space parameters of the parking space. The simulated parking space frame corresponding to one point is used to determine whether there is at least one point of the third point cloud in the simulated parking space frame; if so, the simulated parking space frame is an invalid simulated parking space frame; if not, the simulated parking space frame is valid Simulated parking space frame;
统计单元，用于统计所有有效模拟车位框所对应的点，并根据所述所有有效模拟车位框所对应的点生成空车位的车位框。The statistical unit is configured to count the points corresponding to all valid simulated parking space frames, and generate a parking space frame of an empty parking space according to the points corresponding to all the valid simulated parking space frames.
本发明第三方面提出一种计算机可读存储介质，其上存储有计算机程序，所述计算机程序被处理器执行时实现第一方面所述停车场空车位识别方法。A third aspect of the present invention provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, implements the method for identifying an empty parking space in a parking lot according to the first aspect.
综上，本发明的实施例提出一种停车场空车位识别方法及系统、计算机可读存储介质，其至少具有以下优点：To sum up, the embodiments of the present invention provide a method and system for identifying an empty parking space in a parking lot, and a computer-readable storage medium, which at least have the following advantages:
(1)相比于传统的基于超声波雷达的车位识别方法，本发明的实施例在车位平面二维空间的信息点更多而且在垂直方向上也有大量信息点，大大提高空车位识别的准确率；(1) Compared with the traditional ultrasonic radar-based parking space identification method, the embodiment of the present invention has more information points in the two-dimensional space of the parking space plane and also has a large number of information points in the vertical direction, which greatly improves the accuracy of empty parking space identification. ;
(2)相比于传统的基于地磁传感器的车位识别方法，不需要提前实现对停车区域的整体改造，应用方便；(2) Compared with the traditional parking space identification method based on the geomagnetic sensor, it is not necessary to realize the overall transformation of the parking area in advance, and the application is convenient;
(3)相比传统的基于视觉的车位识别方法，本发明的实施例对空车位的检测不依赖于车位线，提高了自动泊车过程中空车位检测的鲁棒性。(3) Compared with the traditional vision-based parking space identification method, the embodiment of the present invention does not rely on the parking space line to detect the empty parking space, which improves the robustness of the empty parking space detection in the automatic parking process.
本发明的其它特征和优点将在随后的说明书中阐述。Other features and advantages of the present invention will be set forth in the description that follows.
附图说明Description of drawings
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。In order to explain the embodiments of the present invention or the technical solutions in the prior art more clearly, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained according to these drawings without creative efforts.
图1为本发明一实施例中一种停车场空车位识别方法流程图。FIG. 1 is a flowchart of a method for identifying an empty parking space in a parking lot according to an embodiment of the present invention.
图2为本发明一实施例中一种停车场空车位识别系统结构示意图。FIG. 2 is a schematic structural diagram of an empty parking space identification system in a parking lot according to an embodiment of the present invention.
具体实施方式Detailed ways
以下将参考附图详细说明本公开的各种示例性实施例、特征和方面。另外，为了更好的说明本发明，在下文的具体实施例中给出了众多的具体细节。本领域技术人员应当理解，没有某些具体细节，本发明同样可以实施。在一些实例中，对于本领域技术人员熟知的手段未作详细描述，以便于凸显本发明的主旨。Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. In addition, in order to better illustrate the present invention, numerous specific details are given in the following specific embodiments. It will be understood by those skilled in the art that the present invention may be practiced without certain specific details. In some instances, means well known to those skilled in the art have not been described in detail in order not to obscure the subject matter of the present invention.
参阅图1，本发明一实施例提出一种停车场空车位识别方法，包括如下步骤S1～S5：Referring to FIG. 1, an embodiment of the present invention provides a method for identifying an empty parking space in a parking lot, which includes the following steps S1-S5:
步骤S1、获取车辆行驶过程中车载深度相机对车辆单侧方向进行探测获得的深度图像；Step S1, acquiring a depth image obtained by detecting the unilateral direction of the vehicle by the vehicle-mounted depth camera during the driving process of the vehicle;
具体而言，本实施例实施过程中，驾驶员驾驶车辆行驶，设置于车辆一侧的车载深度相机车辆单侧方向进行探测获得的深度图像；Specifically, in the implementation process of this embodiment, the driver drives the vehicle to drive, and the depth image obtained by the vehicle-mounted depth camera disposed on one side of the vehicle detects the one-side direction of the vehicle;
步骤S2、对所述深度图像进行点云转换获得第一点云；Step S2, performing point cloud conversion on the depth image to obtain a first point cloud;
具体而言，所述深度相机的像素例如是320×240，根据所述深度相机的内参，对所述深度图像进行三维坐标转换，转换后获得的第一点云包含点数为76800个，全部点以三维空间坐标(X,Y,Z)表示，转换的参考坐标原点为所述深度相机光心，参考坐标系根据右手法则确立；Specifically, the pixels of the depth camera are, for example, 320×240. According to the internal parameters of the depth camera, three-dimensional coordinate transformation is performed on the depth image. The first point cloud obtained after the transformation contains 76,800 points. It is represented by three-dimensional space coordinates (X, Y, Z), the origin of the converted reference coordinates is the optical center of the depth camera, and the reference coordinate system is established according to the right-hand rule;
步骤S3、按照预设滤波规则对所述第一点云进行滤波获得第二点云；Step S3, filtering the first point cloud according to a preset filtering rule to obtain a second point cloud;
具体而言，步骤S2获得的第一点云为原始点云，其中可能还包含一些无效点或离群点，需要进行滤波获得第二点云；Specifically, the first point cloud obtained in step S2 is the original point cloud, which may also contain some invalid points or outliers, and needs to be filtered to obtain the second point cloud;
示例性地，所述步骤S3，包括：Exemplarily, the step S3 includes:
步骤S31、进行一次滤波，以过滤掉所述第一点云中深度值等于预设深度值的点；Step S31, performing a filter to filter out the point whose depth value is equal to the preset depth value in the first point cloud;
具体而言，所述一次滤波为条件滤波，例如，使用的深度相机有效探测范围是最大6.0米，对于超出探测范围的点则认为深度值为最大深度，深度值对应为三维空间坐标(X,Y,Z)中的Z，因此认为Z等于最大深度的点为无效点，需要通过条件滤波把他们过滤掉，本实施例中条件滤波设定Z大于0且小于5.9的点为有效点；Specifically, the first filtering is conditional filtering. For example, the effective detection range of the depth camera used is a maximum of 6.0 meters. For points beyond the detection range, the depth value is considered to be the maximum depth, and the depth value corresponds to the three-dimensional space coordinates (X, Z in Y, Z), therefore it is considered that the points where Z is equal to the maximum depth are invalid points, and they need to be filtered out by conditional filtering. In the present embodiment, the conditional filtering sets the points where Z is greater than 0 and less than 5.9 as valid points;
步骤S32、进行二次滤波，以过滤掉所述第一点云中的离群点；Step S32, performing secondary filtering to filter out outliers in the first point cloud;
具体而言，所述二次滤波为统计滤波，即通过统计点云中所有点的各自邻近点的数量来辨别离群点，把离群点当作噪点进行剔除处理；Specifically, the secondary filtering is statistical filtering, that is, outliers are identified by counting the number of adjacent points of all points in the point cloud, and the outliers are treated as noise points for elimination;
优选地，本实施例中统计滤波在统计时考虑查询点的邻近点数量为50个，判断是否为离群点的阈值为0.3；Preferably, the statistical filtering in this embodiment considers that the number of adjacent points of the query point is 50, and the threshold for judging whether it is an outlier is 0.3;
步骤S4、对所述第二点云进行下采样获得第三点云；Step S4, downsampling the second point cloud to obtain a third point cloud;
示例性地，所述步骤S4，包括：Exemplarily, the step S4 includes:
把所述第二点云划分成多个立方体，每个立方体中保留1个点，从而获得第三点云；Divide the second point cloud into a plurality of cubes, and retain 1 point in each cube, thereby obtaining a third point cloud;
具体而言，本示例中利用体素格对点云进行下采样，即把点云划分成许多小立方体，每个立方体中保留1个点以不破坏点云分布特征的前提下减少点的数量，降低后期处理的运算复杂度，本示例中体素块设定为0.01×0.01×0.01米的立方体；Specifically, in this example, the voxel grid is used to downsample the point cloud, that is, the point cloud is divided into many small cubes, and one point is reserved in each cube to reduce the number of points without destroying the distribution characteristics of the point cloud. , to reduce the computational complexity of post-processing. In this example, the voxel block is set to a cube of 0.01 × 0.01 × 0.01 meters;
步骤S5、根据所述第三点云以及预设的车位三维空间参数识别所述第三点云中是否存在空车位；若存在空车位，则生成空车位的车位框。Step S5: Identify whether there is an empty parking space in the third point cloud according to the third point cloud and the preset three-dimensional space parameters of the parking space; if there is an empty parking space, generate a parking space frame of the empty parking space.
示例性地，所述步骤S5，包括：Exemplarily, the step S5 includes:
步骤S51、遍历所述第三点云的点云空间中所有点，将每一点作为车位框左上角点，即空车位原点，并根据该车位框左上角点以及所述预设的车位三维空间参数生成每一点所对应的模拟车位框，判断该模拟车位框中是否存在至少一个所述第三点云的点；若是，则该模拟车位框为无效模拟车位框；若否，则该模拟车位框为有效模拟车位框；Step S51, traverse all points in the point cloud space of the third point cloud, take each point as the upper left corner point of the parking space frame, that is, the origin of the empty parking space, and according to the upper left corner point of the parking space frame and the preset three-dimensional space of the parking space The parameters generate a simulated parking space frame corresponding to each point, and determine whether there is at least one point of the third point cloud in the simulated parking space frame; if so, the simulated parking space frame is an invalid simulated parking space frame; if not, the simulated parking space The box is a valid simulated parking space box;
步骤S52、统计所有有效模拟车位框所对应的点，并根据所述所有有效模拟车位框所对应的点生成空车位的车位框；Step S52, count the points corresponding to all the valid simulated parking space frames, and generate a parking space frame of an empty parking space according to the corresponding points of all the valid simulated parking space frames;
具体而言，本示例中基于下采样后的点云计算是否满足车位空间要求；设定车位三维空间参数，例如，长宽高分别为5.0×1.8×1.6米，把车位框的左上角点作为车位原点，先假设车位原点位于某个地方，然后按照车位长宽高生成对应模拟车位框，在该模拟车位框中遍历所有点云，只要模拟车位框中存在至少一个有效点，则认为该模拟车位框假设不成立；通过改变车位原点位置，遍历所述第三点云的点云空间中所有的点，穷举点云空间，找出满足条件的所有车位框，即空车位；Specifically, in this example, whether the down-sampled point cloud meets the requirements of the parking space space is calculated; the three-dimensional space parameters of the parking space are set, for example, the length, width and height are 5.0×1.8×1.6 meters respectively, and the upper left corner of the parking space frame is used as The origin of the parking space, first assume that the origin of the parking space is located somewhere, and then generate a corresponding simulated parking space frame according to the length, width and height of the parking space, and traverse all the point clouds in the simulated parking space frame. As long as there is at least one valid point in the simulated parking space frame, it is considered that the simulation The assumption of the parking space frame does not hold; by changing the position of the origin of the parking space, traverse all the points in the point cloud space of the third point cloud, exhaust the point cloud space, and find all the parking space frames that meet the conditions, that is, empty parking spaces;
例如，假设车位原点的初始坐标(X,Y,Z)为(-1.9,-0.8,0.5)，车位原点Y坐标固定在-0.8，Z坐标固定在0.5，X坐标则从左到右按0.1的步进幅度增加，进行平移迭代计算，边界条件为X坐标初始值的相反数-车位宽度，即1.9-1.8＝0.1，即车位原点X坐标从-1.9平移到0.1，步进长度0.1。最后，把所有满足条件的空车位原点集合成线段，作为有效空车位的位置分布，以线段的形式记录；For example, assuming that the initial coordinates (X, Y, Z) of the origin of the parking space are (-1.9, -0.8, 0.5), the Y coordinate of the origin of the parking space is fixed at -0.8, the Z coordinate is fixed at 0.5, and the X coordinate is 0.1 from left to right The step amplitude increases, and the translation iterative calculation is performed. The boundary condition is the opposite number of the initial value of the X coordinate - the width of the parking space, that is, 1.9-1.8=0.1, that is, the X coordinate of the origin of the parking space is translated from -1.9 to 0.1, and the step length is 0.1. Finally, all the origins of empty parking spaces that meet the conditions are assembled into line segments, which are used as the location distribution of valid empty parking spaces and recorded in the form of line segments;
在一具体实施例中，所述方法还包括：In a specific embodiment, the method further includes:
步骤S6、根据所述第一点云、第二点云、第三点云以及所述车位框分别生成第一显示信息、第二显示信息、第三显示信息、第四显示信息，并将所述第一显示信息、第二显示信息、第三显示信息、第四显示信息发送至车载显示单元进行同步显示。Step S6: Generate first display information, second display information, third display information, and fourth display information according to the first point cloud, the second point cloud, the third point cloud and the parking space frame, respectively, and put all the The first display information, the second display information, the third display information, and the fourth display information are sent to the vehicle-mounted display unit for synchronous display.
具体而言，本实施例中提出了点云可视化，点云可视化首先创建了视窗，然后在车载显示单元的视窗中创建了四个视点，每个视点联动显示，即拖动一个视点变换旋转角度或放大缩小，其他三个视点也伴随发生相同的变换，即同步显示；四个视点分别为所述第一显示信息、第二显示信息、第三显示信息、第四显示信息。其中，包含车位框的第四显示信息，是根据上一步最后生成的有效空车位线段来重新生成车位框，例如假如空车位线段X坐标从-1.9到0.1的话，车位框则变成5.0×2.0×1.6米。Specifically, point cloud visualization is proposed in this embodiment. The point cloud visualization first creates a viewport, and then creates four viewpoints in the viewport of the vehicle-mounted display unit, and each viewpoint is displayed in linkage, that is, dragging a viewpoint to transform the rotation angle Or zoom in and out, the other three viewpoints also undergo the same transformation, that is, synchronous display; the four viewpoints are the first display information, the second display information, the third display information, and the fourth display information, respectively. Among them, the fourth display information including the parking space frame is to regenerate the parking space frame according to the last valid empty parking space line segment generated in the previous step. For example, if the X coordinate of the empty parking space line segment is from -1.9 to 0.1, the parking space frame becomes 5.0×2.0 ×1.6 meters.
本实施例方法通过单个深度相机对车辆行驶过程中的单侧方向进行探测，得到单帧深度图后，先将深度图变换成点云，再对点云进行滤波、下采样等处理，然后通过遍历点云完成空车位的假设检验和空车位本身的位置穷举，在有效距离内精确识别出空车位的位置集合；基于以上描述可知，本发明实施例具有以下优点：The method in this embodiment uses a single depth camera to detect the unilateral direction of the vehicle during driving, and after obtaining a single-frame depth map, the depth map is first transformed into a point cloud, and then the point cloud is filtered and downsampled. Traverse the point cloud to complete the hypothesis test of the empty parking space and the exhaustive position of the empty parking space itself, and accurately identify the position set of the empty parking space within the effective distance; based on the above description, it can be seen that the embodiment of the present invention has the following advantages:
(1)相比于传统的基于超声波雷达的车位识别方法，本发明的实施例在车位平面二维空间的信息点更多而且在垂直方向上也有大量信息点，大大提高空车位识别的准确率；(1) Compared with the traditional ultrasonic radar-based parking space identification method, the embodiment of the present invention has more information points in the two-dimensional space of the parking space plane and also has a large number of information points in the vertical direction, which greatly improves the accuracy of empty parking space identification. ;
(2)相比于传统的基于地磁传感器的车位识别方法，不需要提前实现对停车区域的整体改造，应用方便；(2) Compared with the traditional parking space identification method based on the geomagnetic sensor, it is not necessary to realize the overall transformation of the parking area in advance, and the application is convenient;
(3)相比传统的基于视觉的车位识别方法，本发明的实施例对空车位的检测不依赖于车位线，提高了自动泊车过程中空车位检测的鲁棒性。(3) Compared with the traditional vision-based parking space identification method, the embodiment of the present invention does not rely on the parking space line to detect the empty parking space, which improves the robustness of the empty parking space detection in the automatic parking process.
(4)相比传统的基于深度学习的点云车位识别方法，本发明的实施例使用简单逻辑规则实现车位检测，节省了数据采集处理的工作，降低运算复杂度，从而降低了对硬件的算力要求，也方便针对不同停车场实景进行参数调整迁移，方法逻辑性强有利于功能的实现和问题的排查与修复。(4) Compared with the traditional point cloud parking space identification method based on deep learning, the embodiment of the present invention uses simple logic rules to realize parking space detection, which saves the work of data collection and processing, reduces the computational complexity, and thus reduces the computational complexity of hardware. It is also convenient for parameter adjustment and migration for different parking lot real scenes. The method is logical and conducive to the realization of functions and the troubleshooting and repair of problems.
参阅图2，本发明另一实施例提出一种停车场空车位识别系统，包括：Referring to FIG. 2, another embodiment of the present invention proposes a parking space identification system, including:
图像获取单元1，用于获取车辆行驶过程中车载深度相机对车辆单侧方向进行探测获得的深度图像；The image acquisition unit 1 is used for acquiring the depth image obtained by the vehicle-mounted depth camera detecting the unilateral direction of the vehicle during the driving of the vehicle;
点云转换单元2，用于对所述深度图像进行点云转换获得第一点云；A point
点云过滤单元3，用于按照预设滤波规则对所述第一点云进行滤波获得第二点云；A point
点云采样单元4，用于对所述第二点云进行下采样获得第三点云；以及a point cloud sampling unit 4, configured to downsample the second point cloud to obtain a third point cloud; and
空车位识别单元5，用于根据所述第三点云以及预设的车位三维空间参数识别所述第三点云中是否存在空车位；若存在空车位，则生成空车位的车位框。The empty parking
示例性地，所述点云过滤单元，包括：Exemplarily, the point cloud filtering unit includes:
第一滤波单元，用于进行一次滤波，以过滤掉所述第一点云中深度值等于预设深度值的点；a first filtering unit, configured to perform a filtering to filter out points whose depth value is equal to a preset depth value in the first point cloud;
第二滤波单元，用于进行二次滤波，以过滤掉所述第一点云中的离群点。The second filtering unit is configured to perform secondary filtering to filter out outliers in the first point cloud.
示例性地，所述点云采样单元，具体用于Exemplarily, the point cloud sampling unit is specifically used for
把所述第二点云划分成多个立方体，每个立方体中保留1个点，从而获得第三点云。The second point cloud is divided into a plurality of cubes, and one point is reserved in each cube, thereby obtaining a third point cloud.
示例性地，所述空车位识别单元，具体包括：Exemplarily, the empty parking space identification unit specifically includes:
模拟单元，用于遍历所述第三点云的点云空间中所有点，将每一点作为车位框左上角点，并根据该车位框左上角点以及所述预设的车位三维空间参数生成每一点所对应的模拟车位框，判断该模拟车位框中是否存在至少一个所述第三点云的点；若是，则该模拟车位框为无效模拟车位框；若否，则该模拟车位框为有效模拟车位框；The simulation unit is used to traverse all the points in the point cloud space of the third point cloud, take each point as the upper left corner point of the parking space frame, and generate each point according to the upper left corner point of the parking space frame and the preset three-dimensional space parameters of the parking space. The simulated parking space frame corresponding to one point is used to determine whether there is at least one point of the third point cloud in the simulated parking space frame; if so, the simulated parking space frame is an invalid simulated parking space frame; if not, the simulated parking space frame is valid Simulated parking space frame;
统计单元，用于统计所有有效模拟车位框所对应的点，并根据所述所有有效模拟车位框所对应的点生成空车位的车位框。The statistical unit is configured to count the points corresponding to all valid simulated parking space frames, and generate a parking space frame of an empty parking space according to the points corresponding to all the valid simulated parking space frames.
以上所描述的系统实施例仅仅是示意性的，其中所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。The system embodiments described above are only illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in One place, or it can be distributed over multiple network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution in this embodiment.
需说明的是，上述实施例所述系统与上述实施例所述方法对应，因此，上述实施例所述系统未详述部分可以参阅上述实施例所述方法的内容得到，此处不再赘述。It should be noted that the system described in the foregoing embodiment corresponds to the method described in the foregoing embodiment. Therefore, the undescribed part of the system described in the foregoing embodiment can be obtained by referring to the content of the method described in the foregoing embodiment, which will not be repeated here.
并且，上述实施例所述停车场空车位识别系统若以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读取存储介质中。Moreover, if the parking lot vacant space identification system described in the above embodiment is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium.
本发明另一实施例提出一种计算机可读存储介质，其上存储有计算机程序，所述计算机程序被处理器执行时实现上述实施例所述停车场空车位识别方法。Another embodiment of the present invention provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, implements the method for identifying an empty parking space in a parking lot according to the above embodiment.
具体而言，所述计算机可读存储介质可以包括：能够携带所述计算机程序代码的任何实体或装置、记录介质、U盘、移动硬盘、磁碟、光盘、计算机存储器、只读存储器(R值OM，R值ead-Only MemoR值y)、随机存取存储器(R值AM，R值andom Access MemoR值y)、电载波信号、电信信号以及软件分发介质等。Specifically, the computer-readable storage medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, removable hard disk, magnetic disk, optical disk, computer memory, read-only memory (R-value) OM, R-value ead-Only MemoR-value y), random access memory (R-value AM, R-value anddom Access MemoR-value y), electrical carrier signal, telecommunication signal, and software distribution medium, etc.
以上已经描述了本发明的各实施例，上述说明是示例性的，并非穷尽性的，并且也不限于所披露的各实施例。在不偏离所说明的各实施例的范围和精神的情况下，对于本技术领域的普通技术人员来说许多修改和变更都是显而易见的。本文中所用术语的选择，旨在最好地解释各实施例的原理、实际应用或对市场中的技术改进，或者使本技术领域的其它普通技术人员能理解本文披露的各实施例。Various embodiments of the present invention have been described above, and the foregoing descriptions are exemplary, not exhaustive, and not limiting of the disclosed embodiments. Numerous modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terminology used herein was chosen to best explain the principles of the embodiments, the practical application or technical improvement in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.
Claims (10)
Translated fromChinesePriority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110088091.9ACN114783172B (en) | 2021-01-22 | 2021-01-22 | A parking lot vacant parking space identification method and system, and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110088091.9ACN114783172B (en) | 2021-01-22 | 2021-01-22 | A parking lot vacant parking space identification method and system, and computer readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114783172Atrue CN114783172A (en) | 2022-07-22 |
| CN114783172B CN114783172B (en) | 2024-05-17 |
Family
ID=82407752
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110088091.9AActiveCN114783172B (en) | 2021-01-22 | 2021-01-22 | A parking lot vacant parking space identification method and system, and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114783172B (en) |
Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2436577A2 (en)* | 2010-09-30 | 2012-04-04 | Valeo Schalter und Sensoren GmbH | Device and method for detecting free parking spots |
| US20170017848A1 (en)* | 2015-07-17 | 2017-01-19 | Magna Electronics Inc. | Vehicle parking assist system with vision-based parking space detection |
| US20180018870A1 (en)* | 2016-07-15 | 2018-01-18 | Intel Corporation | Parking management system and method |
| JP2018116014A (en)* | 2017-01-20 | 2018-07-26 | パイオニア株式会社 | Own vehicle position estimation device, control method, program, and storage medium |
| CN109766404A (en)* | 2019-02-12 | 2019-05-17 | 湖北亿咖通科技有限公司 | Points cloud processing method, apparatus and computer readable storage medium |
| CN109817018A (en)* | 2019-02-20 | 2019-05-28 | 东软睿驰汽车技术（沈阳）有限公司 | A kind of automatic parking method and relevant apparatus |
| CN109871745A (en)* | 2018-12-29 | 2019-06-11 | 百度在线网络技术（北京）有限公司 | Identify method, system and the vehicle of empty parking space |
| CN109872355A (en)* | 2019-01-25 | 2019-06-11 | 哈工大机器人（合肥）国际创新研究院 | A kind of shortest distance acquisition methods and device based on depth camera |
| CN110211420A (en)* | 2019-06-25 | 2019-09-06 | 重庆长安汽车股份有限公司 | Judge current scene whether be parking lot method, system and computer readable storage medium |
| CN110766979A (en)* | 2019-11-13 | 2020-02-07 | 奥特酷智能科技(南京)有限公司 | Parking space detection method for automatic driving vehicle |
| CN110807412A (en)* | 2019-10-30 | 2020-02-18 | 驭势科技（北京）有限公司 | Vehicle laser positioning method, vehicle-mounted equipment and storage medium |
| KR20200020186A (en)* | 2018-08-16 | 2020-02-26 | 주식회사 다산네트웍스 | Method for parking management of autonomous vehicle |
| CN110889974A (en)* | 2018-09-11 | 2020-03-17 | 广州汽车集团股份有限公司 | Intelligent parking space identification method and device and automobile |
| CN111415526A (en)* | 2020-03-27 | 2020-07-14 | 北京百度网讯科技有限公司 | Method, device, electronic device and storage medium for obtaining parking space occupancy status |
| US20200258385A1 (en)* | 2019-02-11 | 2020-08-13 | Byton North America Corporation | Advanced detection of parking spot for vehicle |
| CN112009462A (en)* | 2020-08-10 | 2020-12-01 | 广州汽车集团股份有限公司 | Forward automatic parking method and device |
- 2021
- 2021-01-22CNCN202110088091.9Apatent/CN114783172B/enactiveActive
Patent Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2436577A2 (en)* | 2010-09-30 | 2012-04-04 | Valeo Schalter und Sensoren GmbH | Device and method for detecting free parking spots |
| US20170017848A1 (en)* | 2015-07-17 | 2017-01-19 | Magna Electronics Inc. | Vehicle parking assist system with vision-based parking space detection |
| US20180018870A1 (en)* | 2016-07-15 | 2018-01-18 | Intel Corporation | Parking management system and method |
| JP2018116014A (en)* | 2017-01-20 | 2018-07-26 | パイオニア株式会社 | Own vehicle position estimation device, control method, program, and storage medium |
| KR20200020186A (en)* | 2018-08-16 | 2020-02-26 | 주식회사 다산네트웍스 | Method for parking management of autonomous vehicle |
| CN110889974A (en)* | 2018-09-11 | 2020-03-17 | 广州汽车集团股份有限公司 | Intelligent parking space identification method and device and automobile |
| CN109871745A (en)* | 2018-12-29 | 2019-06-11 | 百度在线网络技术（北京）有限公司 | Identify method, system and the vehicle of empty parking space |
| CN109872355A (en)* | 2019-01-25 | 2019-06-11 | 哈工大机器人（合肥）国际创新研究院 | A kind of shortest distance acquisition methods and device based on depth camera |
| US20200258385A1 (en)* | 2019-02-11 | 2020-08-13 | Byton North America Corporation | Advanced detection of parking spot for vehicle |
| CN109766404A (en)* | 2019-02-12 | 2019-05-17 | 湖北亿咖通科技有限公司 | Points cloud processing method, apparatus and computer readable storage medium |
| CN109817018A (en)* | 2019-02-20 | 2019-05-28 | 东软睿驰汽车技术（沈阳）有限公司 | A kind of automatic parking method and relevant apparatus |
| CN110211420A (en)* | 2019-06-25 | 2019-09-06 | 重庆长安汽车股份有限公司 | Judge current scene whether be parking lot method, system and computer readable storage medium |
| CN110807412A (en)* | 2019-10-30 | 2020-02-18 | 驭势科技（北京）有限公司 | Vehicle laser positioning method, vehicle-mounted equipment and storage medium |
| CN110766979A (en)* | 2019-11-13 | 2020-02-07 | 奥特酷智能科技(南京)有限公司 | Parking space detection method for automatic driving vehicle |
| CN111415526A (en)* | 2020-03-27 | 2020-07-14 | 北京百度网讯科技有限公司 | Method, device, electronic device and storage medium for obtaining parking space occupancy status |
| CN112009462A (en)* | 2020-08-10 | 2020-12-01 | 广州汽车集团股份有限公司 | Forward automatic parking method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114783172B (en) | 2024-05-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110032949B (en) | Target detection and positioning method based on lightweight convolutional neural network | |
| CN110569704B (en) | A Multi-strategy Adaptive Lane Line Detection Method Based on Stereo Vision | |
| CN111209770B (en) | Lane line recognition method and device | |
| CN105225482B (en) | Vehicle detecting system and method based on binocular stereo vision | |
| KR20160123668A (en) | Device and method for recognition of obstacles and parking slots for unmanned autonomous parking | |
| CN102629326A (en) | Lane line detection method based on monocular vision | |
| CN112446227A (en) | Object detection method, device and equipment | |
| CN109997148A (en) | Information processing apparatus, imaging apparatus, device control system, moving object, information processing method, and computer-readable recording medium | |
| CN114937081B (en) | Method and device for networked vehicle position estimation based on independent non-uniform incremental sampling | |
| CN117274036A (en) | Parking scene detection method based on multi-view and time sequence fusion | |
| CN116052120A (en) | Excavator night object detection method based on image enhancement and multi-sensor fusion | |
| CN118898824B (en) | Radar-vision fusion target perception method, system, device, processor and computer-readable storage medium based on hybrid strategy | |
| CN114445648A (en) | Obstacle identification method, device and storage medium | |
| CN111563457A (en) | Road scene segmentation method for unmanned automobile | |
| CN115546027B (en) | Image suture line determination method, device and storage medium | |
| CN118149797B (en) | Grid map construction method, device, computer equipment and storage medium | |
| CN109492454B (en) | Object identification method and device | |
| CN111208530B (en) | Positioning layer generation method, device, high-precision map and equipment | |
| CN115115535A (en) | Depth map denoising method, device, medium and device | |
| CN108647579B (en) | Obstacle detection method and device and terminal | |
| CN114757959A (en) | Obstacle and cliff detection method, device, readable storage medium and robot | |
| CN116704151A (en) | Three-dimensional reconstruction method and device, and vehicle, equipment and medium based on three-dimensional reconstruction method and device | |
| CN117372680B (en) | Target detection method based on fusion of binocular camera and laser radar | |
| CN114783172B (en) | A parking lot vacant parking space identification method and system, and computer readable storage medium | |
| CN118823732A (en) | Lane line recognition method, device, vehicle and readable storage medium |
Legal Events
| 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |