CN116363692A - Human body shielding detection method, computer equipment and storage medium - Google Patents
Human body shielding detection method, computer equipment and storage mediumDownload PDFInfo
- Publication number
- CN116363692A CN116363692ACN202310090877.3ACN202310090877ACN116363692ACN 116363692 ACN116363692 ACN 116363692ACN 202310090877 ACN202310090877 ACN 202310090877ACN 116363692 ACN116363692 ACN 116363692A
- Authority
- CN
- China
- Prior art keywords
- area
- human body
- image
- pixel value
- region
- 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.)
- Pending
Links
Images
Classifications
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/103—Static body considered as a whole, e.g. static pedestrian or occupant recognition
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30168—Image quality inspection
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30196—Human being; Person
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Geometry (AREA)
- Image Analysis (AREA)
Abstract
Description
Translated fromChinese技术领域technical field
本发明涉及图像处理技术领域，尤其涉及一种人体遮挡检测方法、计算机设备和存储介质。The invention relates to the technical field of image processing, in particular to a human body occlusion detection method, computer equipment and a storage medium.
背景技术Background technique
现有的遮挡检测基本上都是基于将行人检测后的图像送入图像分类网络进行遮挡分类或通过预测行人关键点坐标的可见性通过判断关键点坐标置信度和设置的可能遮挡阈值高低来间接判断该行人是否存在遮挡。The existing occlusion detection is basically based on sending the detected images of pedestrians to the image classification network for occlusion classification or by predicting the visibility of pedestrian key point coordinates and indirectly by judging the confidence of key point coordinates and setting the possible occlusion threshold. Determine whether the pedestrian is occluded.
基于关键点坐标置信度的方法对遮挡判断的可解释性不强且随着光照环境的改变容易误判，且不同环境可能需要设置不同的置信度阈值。图像分类的遮挡检测往往通过对分类结果的类别分数高低进行是否遮挡判断，其缺乏对遮挡判断的可解释性。The method based on the confidence of key point coordinates is not very interpretable for occlusion judgment and is prone to misjudgment as the lighting environment changes, and different environments may need to set different confidence thresholds. The occlusion detection of image classification often judges whether it is occluded or not based on the category score of the classification result, which lacks the interpretability of occlusion judgment.
发明内容Contents of the invention
基于此，有必要针对上述问题，提出了一种人体遮挡检测方法、计算机设备和存储介质，能够提升人体遮挡检测的可靠性和准确性。Based on this, it is necessary to address the above problems and propose a human body occlusion detection method, computer equipment and storage media, which can improve the reliability and accuracy of human body occlusion detection.
一种人体遮挡检测方法，包括：A human body occlusion detection method, comprising:
获取待处理图像，裁减出所述待处理图像中的人体图像；Acquiring the image to be processed, and cutting out the human body image in the image to be processed;
将所述人体图像输入神经网络模型，获取所述人体图像的关键点位置和语义分割结果，基于所述关键点位置和所述语义分割结果获取所述人体图像的人体区域分割图像，所述人体区域分割图像包括多个分割区域，每个所述分割区域具有对应的像素值；Inputting the human body image into a neural network model, obtaining key point positions and semantic segmentation results of the human body image, and obtaining human body region segmentation images of the human body image based on the key point positions and the semantic segmentation results, the human body The region segmentation image includes a plurality of segmented regions, each of which has a corresponding pixel value;
获取所述关键点位置在所述人体区域分割图像中对应的点的图像像素值，判断所述图像像素值是否为有效像素值；Acquiring the image pixel value of the point corresponding to the key point position in the human body region segmentation image, and judging whether the image pixel value is a valid pixel value;
若所有关键点位置对应的点的图像像素值均为有效像素值，则获取各个所述分割区域的有效像素值区域面积和区域连通性，基于所述有效像素值区域面积和/或所述区域连通性判断各个所述分割区域是否被遮挡。If the image pixel values of the points corresponding to all key point positions are effective pixel values, then obtain the effective pixel value area and area connectivity of each of the segmented areas, based on the effective pixel value area and/or the area Connectivity judges whether each of the segmented regions is blocked.
其中，所述获取各个所述分割区域的有效像素值区域面积和区域连通性的步骤，包括：Wherein, the step of obtaining the effective pixel value region area and region connectivity of each of the segmented regions includes:
判断每个所述分割区域的有效像素值区域面积是否低于对应的预设面积阈值，若所述分割区域的有效像素值区域面积低于所述对应的预设面积阈值，则判定所述分割区域存在遮挡情况。judging whether the effective pixel value region area of each of the segmented regions is lower than the corresponding preset area threshold, and if the effective pixel value region area of the segmented region is lower than the corresponding preset area threshold value, then determine that the segmented region There is occlusion in the area.
其中，所述分割区域包括头部区域、上半身区域和下半身区域；Wherein, the segmented area includes a head area, an upper body area and a lower body area;
所述判断所述分割区域的有效像素值区域面积是否低于预设面积阈值的步骤之后，包括：After the step of judging whether the effective pixel value region area of the segmented region is lower than the preset area threshold, it includes:
若所述分割区域的有效像素值区域面积不低于的所述对应的预设面积阈值，将所述下半身区域和所述上半身区域的有效像素值区域面积相比，若比值低于预设比值阈值，则判定所述下半身区域存在遮挡；If the effective pixel value region area of the segmented region is not lower than the corresponding preset area threshold, compare the effective pixel value region areas of the lower body region and the upper body region, and if the ratio is lower than the preset ratio threshold, it is determined that the lower body area is occluded;
若比值不低于预设比值阈值，判断所述上半身区域和所述下半身区域之间的区域是否均为有效像素值，若所述上半身区域和所述下半身区域之间的区域不是均为有效像素值，则判定人体中间部分存在遮挡，且判定所述人体图像的质量不满足预设要求。If the ratio is not lower than the preset ratio threshold, determine whether the area between the upper body area and the lower body area is a valid pixel value, if the area between the upper body area and the lower body area is not all valid pixels value, it is determined that there is occlusion in the middle part of the human body, and it is determined that the quality of the human body image does not meet the preset requirements.
其中，所述基于所述有效像素值区域面积和所述区域连通性判断各个所述分割区域是否被遮挡的步骤，包括：Wherein, the step of judging whether each of the segmented regions is blocked based on the effective pixel value region area and the region connectivity includes:
判断是否存在所述分割区域的有效像素值区域面积为0，若存在所述分割区域的有效像素值区域面积为0，则判定所述人体图像的质量不满足预设要求。Judging whether there is an effective pixel value area of the segmented area is 0, and if there is an effective pixel value area of the segmented area is 0, it is determined that the quality of the human body image does not meet the preset requirements.
其中，所述基于所述有效像素值区域面积和所述区域连通性判断各个所述分割区域是否被遮挡的步骤，包括：Wherein, the step of judging whether each of the segmented regions is blocked based on the effective pixel value region area and the region connectivity includes:
获取每个所述分割区域中的轮廓的数量，若所述分割区域中的轮廓的数量大于或等于2，则判定所述分割区域存在遮挡。The number of contours in each segmented area is obtained, and if the number of contours in the segmented area is greater than or equal to 2, it is determined that the segmented area is occluded.
其中，所述基于所述有效像素值区域面积和所述区域连通性判断各个所述分割区域是否被遮挡的步骤，包括：Wherein, the step of judging whether each of the segmented regions is blocked based on the effective pixel value region area and the region connectivity includes:
若所述分割区域中的轮廓的数量大于或等于3，则判定所述人体图像的质量不满足预设要求。If the number of contours in the segmented area is greater than or equal to 3, it is determined that the quality of the human body image does not meet a preset requirement.
其中，所述判断所述图像像素值是否为有效像素值的步骤之后，包括：Wherein, after the step of judging whether the image pixel value is a valid pixel value, it includes:
若存在所述关键点位置的图像像素值不是有效像素值，则判定所述图像像素值对应的关键点位置被遮挡。If the image pixel value at the key point position is not a valid pixel value, it is determined that the key point position corresponding to the image pixel value is blocked.
其中，所述判断所述图像像素值是否为有效像素值的步骤之后，包括：Wherein, after the step of judging whether the image pixel value is a valid pixel value, it includes:
计算被遮挡的关键点位置的数量，若所述数量超过预设数量阈值，则判定所述人体图像的质量不满足预设要求。The number of blocked key point positions is calculated, and if the number exceeds a preset number threshold, it is determined that the quality of the human body image does not meet the preset requirement.
一种计算机设备，包括存储器和处理器，所述存储器存储有计算机程序，所述计算机程序被所述处理器执行时，使得所述处理器执行如上所述的步骤。A computer device includes a memory and a processor, the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps described above.
一种计算机可读存储介质，存储有计算机程序，所述计算机程序被处理器执行时，使得所述处理器执行如上所述的步骤。A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor executes the above-mentioned steps.
采用本发明实施例，具有如下有益效果：Adopting the embodiment of the present invention has the following beneficial effects:
获取所述人体图像的关键点位置和语义分割结果，基于所述关键点位置和所述语义分割结果获取所述人体图像的人体区域分割图像，基于每个分割区域的所述有效像素值区域面积和/或所述区域连通性判断各个所述分割区域是否被遮挡，将关键点检测及图像分割方法结合进行遮挡检测，舍弃了关键点置信度而结合语义分割结果进行遮挡判断，不需要根据环境调整阈值，对环境鲁棒性高，检测结果准确且可靠。Acquiring key point positions and semantic segmentation results of the human body image, obtaining human body region segmentation images of the human body image based on the key point positions and the semantic segmentation results, and obtaining the effective pixel value region area of each segmented region And/or the regional connectivity to determine whether each of the segmented regions is occluded, the key point detection and image segmentation method are combined to perform occlusion detection, the key point confidence is discarded and the occlusion judgment is performed in combination with the semantic segmentation results, no need to rely on the environment Adjusting the threshold has high robustness to the environment, and the detection results are accurate and reliable.
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 more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the 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. Those skilled in the art can also obtain other drawings based on these drawings without creative work.
其中：in:
图1是本发明提供的人体遮挡检测方法的第一实施例的流程示意图；Fig. 1 is a schematic flow chart of the first embodiment of the human body occlusion detection method provided by the present invention;
图2是本发明提供的关键点位置的示意图；Fig. 2 is a schematic diagram of key point positions provided by the present invention;
图3是本发明提供的神经网络的结构示意图；Fig. 3 is the structural representation of the neural network provided by the present invention;
图4是本发明提供的人体区域分割图像的示意图；Fig. 4 is a schematic diagram of a human body region segmentation image provided by the present invention;
图5是将关键点位置在人体区域分割图像中对应的点的图像示意图；Fig. 5 is a schematic diagram of an image of a point corresponding to a key point position in a human body region segmentation image;
图6是本发明提供的人体遮挡检测方法的第二实施例的流程示意图；Fig. 6 is a schematic flow chart of the second embodiment of the human body occlusion detection method provided by the present invention;
图7是本发明提供的人体遮挡检测方法的第三实施例的流程示意图；Fig. 7 is a schematic flowchart of the third embodiment of the human body occlusion detection method provided by the present invention;
图8是本发明提供的人体遮挡检测结果的示意图；Fig. 8 is a schematic diagram of human body occlusion detection results provided by the present invention;
图9是本发明提供的计算机设备的一实施例的结构示意图；FIG. 9 is a schematic structural diagram of an embodiment of a computer device provided by the present invention;
图10是本发明提供的计算机可读存储介质的一实施例的结构示意图。Fig. 10 is a schematic structural diagram of an embodiment of a computer-readable storage medium provided by the present invention.
具体实施方式Detailed ways
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.
请结合参阅图1，图1是本发明提供的人体遮挡检测方法的第一实施例的流程示意图。本发明提供的人体遮挡检测方法包括如下步骤：Please refer to FIG. 1 in conjunction with FIG. 1 . FIG. 1 is a schematic flowchart of a first embodiment of a method for detecting human body occlusion provided by the present invention. The human body occlusion detection method provided by the present invention comprises the following steps:
S101：获取待处理图像，裁减出待处理图像中的人体图像。S101: Acquire an image to be processed, and cut out a human body image in the image to be processed.
在一个具体的实施场景中，获取待处理图像，待处理图像可以是通过对视频进行采样获取的某一帧的图像，也可以是拍摄终端进行实时拍摄获取的图像，或者是网络上下载的图像，还可以是用户上传的图像。将待处理图像输入检测模型，获取待处理图像中的人体轮廓，沿着人体轮廓进行裁剪，获取待处理图像中的人体图像。当待处理图像中存在多个人物时，针对每一个人物的人体轮廓进行裁剪，获取多个人体图像，一个人体图像对应一个人物，分别对每个人体图像进行人体遮挡检测，或者根据预设标准从多个人物中选出目标人物，获取目标人物的人体图像。在获取人体图像后需要对人体图像进行缩放处理，以确保人体图像的大小基本一致，方便进行后续检测处理。In a specific implementation scenario, the image to be processed is obtained. The image to be processed can be an image of a certain frame obtained by sampling a video, or an image obtained by a shooting terminal in real-time shooting, or an image downloaded from the network , which can also be a user-uploaded image. Input the image to be processed into the detection model, obtain the outline of the human body in the image to be processed, cut along the outline of the human body, and obtain the image of the human body in the image to be processed. When there are multiple people in the image to be processed, the human body contour of each person is cropped to obtain multiple human body images, one human body image corresponds to one person, and human body occlusion detection is performed on each human body image, or according to preset standards A target person is selected from multiple persons, and a human body image of the target person is acquired. After acquiring the human body image, it is necessary to perform scaling processing on the human body image to ensure that the size of the human body image is basically the same, which is convenient for subsequent detection processing.
S102：将人体图像输入神经网络模型，获取人体图像的关键点位置和语义分割结果，基于关键点位置和语义分割结果获取人体图像的人体区域分割图像，人体区域分割图像包括多个分割区域，每个分割区域具有对应的像素值。S102: Input the human body image into the neural network model, obtain key point positions and semantic segmentation results of the human body image, and obtain human body region segmentation images of the human body image based on the key point positions and semantic segmentation results. The human body region segmentation image includes multiple segmented regions, each segmented regions have corresponding pixel values.
在一个具体的实施场景中，将人体图像输入神经网络模型，该神经网络模型能够获取人体图像的关键点位置，关键点位置采用COCO数据集对行人关键点的定义方式，由人体17个关键点组成，其中0-鼻子，1-左眼，2-右眼，3-左耳，4-右耳，5-左肩，6-右肩，7-左肘，8-右肘，9-左腕，10-右腕，11-左臀，12-右臀，13-左膝，14-右膝，15-左踝，16-右踝，请结合参阅图2，图2是本发明提供的关键点位置的示意图。该神经网络模型还能够获取人体图像的语义信息，并基于语义信息对人体图像的每个像素点赋予对应的像素值，具体的像素值可以由用户根据实际需求设定，能够让不同的语义信息的点区别显示即可，在本实施场景中，将上半身区域语义标签像素值设为1，下半身区域语义标签像素值设为2，人头区域语义标签像素值设为3，背景区域语义标签像素值设为0，其中，遮挡人物的物体被定义为归于背景类。In a specific implementation scenario, the human body image is input into the neural network model, and the neural network model can obtain the key point position of the human body image. Composition, where 0-nose, 1-left eye, 2-right eye, 3-left ear, 4-right ear, 5-left shoulder, 6-right shoulder, 7-left elbow, 8-right elbow, 9-left wrist, 10-right wrist, 11-left buttock, 12-right buttock, 13-left knee, 14-right knee, 15-left ankle, 16-right ankle, please refer to Fig. 2 in conjunction with Fig. 2, which is the key point position provided by the present invention schematic diagram. The neural network model can also obtain the semantic information of the human body image, and assign a corresponding pixel value to each pixel of the human body image based on the semantic information. The specific pixel value can be set by the user according to actual needs, which can make different semantic information In this implementation scenario, set the semantic label pixel value of the upper body area to 1, set the semantic label pixel value of the lower body area to 2, set the semantic label pixel value of the head area to 3, and set the semantic label pixel value of the background area to Set to 0, where objects that occlude people are defined as belonging to the background class.
请结合参阅图3，图3是本发明提供的神经网络的结构示意图。神经网络模型为基于深度学习的人体关键点检测及分割的多任务模型，同时对人物进行关键点坐标预测及人物像素语义预测。复用神经网络模型浅层的卷积网络部分，其中一个分支用于预测行人关键点坐标，另一个分支用于预测行人图像语义类别，由此组成多任务模型。该神经网络模型的输入为人体图像，经过浅层的权重信息共享的卷积层对输入的人体图像进行纹理、结构等信息的特征提取，相对于两个单独的关键点检测和人体分割任务来说，这一步减少了权重信息共享的卷积层的计算量。通过对两个分支的输出进行处理分别获取当前输入的人体图像对应的关键点位置和语义分割结果。Please refer to FIG. 3 in conjunction with FIG. 3 , which is a schematic structural diagram of the neural network provided by the present invention. The neural network model is a multi-task model for human key point detection and segmentation based on deep learning. At the same time, it performs key point coordinate prediction and character pixel semantic prediction for characters. The convolutional network part of the shallow layer of the neural network model is reused, one branch is used to predict the coordinates of pedestrian key points, and the other branch is used to predict the semantic category of pedestrian images, thus forming a multi-task model. The input of the neural network model is a human body image, and the feature extraction of texture, structure and other information is performed on the input human body image through a convolutional layer sharing shallow weight information. Compared with two separate key point detection and human body segmentation tasks That said, this step reduces the amount of computation in the convolutional layers where weight information is shared. The key point positions and semantic segmentation results corresponding to the currently input human body image are obtained by processing the outputs of the two branches.
同时结合关键点位置和语义分割结果获取人体图像的人体区域分割图像，人体区域分割图像包括多个分割区域，在本实施场景中，多个分割区域包括头部区域、上半身区域、下半身区域。例如，包括关键点0、1、2、3、4，且位于关键点5和6之上的区域属于头部区域，包括关键点5、6、7、8、9、10，且位于关键点11和12之上的区域属于上半身区域，包括关键点11、12、13、13、14、16的区域属于下半身区域，再根据语义分割结果对于头部区域、上半身区域、下半身区域的分界进行调整，获取人体区域分割图像。请结合参阅图4，图4是本发明提供的人体区域分割图像的示意图。获取的人体区域分割图像包括多个分割区域，每个区域具有对应的像素值，例如一个点原本的语义分割结果为1，结合关键点位置，该点属于上半身区域，则该点的像素值为1。再例如一个点的原本的语义分割结果为3，结合关键点位置，该点属于上半身区域，则该点的像素值调整为1。对于语义分割结果为0的点，不进行像素值的调整。At the same time, the human body region segmentation image of the human body image is obtained by combining the key point positions and the semantic segmentation results. The human body region segmentation image includes multiple segmentation regions. In this implementation scenario, the multiple segmentation regions include the head region, upper body region, and lower body region. For example, the area that includes
S103：获取关键点位置在人体区域分割图像中对应的点的图像像素值，判断图像像素值是否为有效像素值，若是，执行步骤S104。S103: Obtain the image pixel value of the point corresponding to the key point position in the human body region segmentation image, and determine whether the image pixel value is a valid pixel value, and if so, perform step S104.
在一个具体的实施场景中，获取关键点位置在人体区域分割图像中对应的点的图像像素值，判断图像像素值是否为有效像素值，有效像素值在本实施场景中为非0像素值，在其他实施场景中，可以是语义信息不是背景的点在步骤S102中被赋予的像素值。获取关键点位置在人体区域分割图像中的图像像素值，请结合参阅图5，图5是将关键点位置在人体区域分割图像中对应的点的图像示意图。若关键点位置对应的点在人体区域分割图像中的图像像素值为0，则表示该关键点位置被遮挡，若关键点位置对应的点在人体区域分割图像中的图像像素值不为0，则表示该关键点位置没有被遮挡，但是并不代表该关键点所在的分割区域没有被遮挡。In a specific implementation scenario, the image pixel value of the point corresponding to the key point position in the human body region segmentation image is obtained, and it is judged whether the image pixel value is an effective pixel value, and the effective pixel value is a non-zero pixel value in this implementation scenario, In other implementation scenarios, it may be the pixel value assigned in step S102 to the point whose semantic information is not the background. To obtain image pixel values of key point positions in the human body region segmentation image, please refer to FIG. 5 , which is an image schematic diagram of key point positions corresponding to points in the human body region segmentation image. If the image pixel value of the point corresponding to the key point position in the human body region segmentation image is 0, it means that the key point position is blocked; if the image pixel value of the point corresponding to the key point position in the human body region segmentation image is not 0, It means that the position of the key point is not occluded, but it does not mean that the segmented area where the key point is located is not occluded.
S104：获取各个分割区域的有效像素值区域面积和区域连通性，基于有效像素值区域面积和/或区域连通性判断各个分割区域是否被遮挡。S104: Obtain the effective pixel value region area and regional connectivity of each segmented region, and determine whether each segmented region is blocked based on the effective pixel value region area and/or region connectivity.
在一个具体的实施场景中，所有关键点位置对应的点的图像像素值均为有效像素值，则表示所有关键点位置不存在遮挡情况，但是还需要排除各个分割区域在关键点位置之外的其余位置是否存在遮挡情况。获取各个分割区域的有效像素值区域面积，以及各个分割区域中区域连通性，也就是统计每个分割区域是否存在多个连通域，基于有效像素值区域面积和/或区域连通性判断各个分割区域是否被遮挡。In a specific implementation scenario, if the image pixel values of the points corresponding to all key point positions are effective pixel values, it means that there is no occlusion at all key point positions, but it is also necessary to exclude the segmented areas outside the key point positions. Whether there is occlusion in other positions. Obtain the effective pixel value area of each segmented area, and the regional connectivity in each segmented area, that is, count whether there are multiple connected domains in each segmented area, and judge each segmented area based on the effective pixel value area and/or regional connectivity Whether it is blocked.
由于分割区域对应的人体部位是已知的，那么分割区域的面积范围也是已知的，例如人的头部的大小是存在一个已知范围的(可以通过大数据采样获取)，则分割区域头部区域的面积就存在一个最小的预设面积阈值，同理，上半身区域和下半身区域都存在各自的预设面积阈值。统计每个分割区域的有效像素值对应的区域的面积，在本实施场景中，是非零像素值对应的面积，也就是统计未被遮挡的区域的面积。将各个分割区域的有效像素值对应的区域的面积和各自的预设面积阈值进行对比，若该分割区域的有效像素值对应的区域的面积低于对应的预设面积阈值，则表示该分割区域有效像素值对应的区域的面积过小，可以判定该分割区域存在被遮挡。Since the body part corresponding to the segmented area is known, the area range of the segmented area is also known. For example, there is a known range of the size of the human head (which can be obtained through large data sampling), then the segmented area head There is a minimum preset area threshold for the area of the upper body area. Similarly, there are respective preset area thresholds for the upper body area and the lower body area. The area corresponding to the effective pixel value of each segmented area is counted. In this implementation scenario, it is the area corresponding to the non-zero pixel value, that is, the area of the unoccluded area is counted. Comparing the area of the area corresponding to the effective pixel value of each segmented area with the respective preset area threshold, if the area of the area corresponding to the effective pixel value of the segmented area is lower than the corresponding preset area threshold, it indicates that the segmented area If the area of the region corresponding to the effective pixel value is too small, it can be determined that the segmented region is occluded.
在其他实施场景中，若该分割区域的有效像素值对应的区域的面积不低于各自的预设面积阈值，则表示该分割区域可能不存在遮挡，需要进行进一步判断是否存在遮挡情况。在本实施场景中，分割区域包括上半身区域和下半身区域，上半身和下半身的面积比例可以是已知的，例如通过大数据采集处理获取。将人体图像分割图像，下半身区域和上半身区域的有效像素值区域面积相比，判断获得的比值是否低于对应的预设比值阈值(例如，0.15)，若比值低于预设比值阈值，则表示下半身区域的有效像素值区域较小，存在部分区域被遮挡住的情况，判定下半身区域存在遮挡。In other implementation scenarios, if the area of the area corresponding to the effective pixel value of the segmented area is not lower than the respective preset area threshold, it means that there may be no occlusion in the segmented area, and it is necessary to further determine whether there is occlusion. In this implementation scenario, the segmented area includes an upper body area and a lower body area, and the area ratio of the upper body and the lower body may be known, for example, acquired through big data collection and processing. Segment the human body image, compare the effective pixel value area of the lower body area and the upper body area, and judge whether the obtained ratio is lower than the corresponding preset ratio threshold (for example, 0.15). If the ratio is lower than the preset ratio threshold, it means The effective pixel value area of the lower body area is small, and some areas are covered, so it is determined that the lower body area is occluded.
在其他实施场景中，若上半身区域和下半身区域均不存在遮挡，则需要判断是否上半身区域和下半身区域之间的区域存在遮挡，例如，本实施场景中，可以判断上半身区域和下半身区域之间的区域是否存在图像像素值为0的点，也就是判断该区域的点的图像像素值是否均为有效像素值，若该区域存在图像像素值为0的点，也就是该区域的点的图像像素值不均为有效像素值，则判定上半身区域和下半身区域之间存在遮挡。In other implementation scenarios, if there is no occlusion in the upper body area and the lower body area, it is necessary to determine whether there is occlusion in the area between the upper body area and the lower body area. Whether there is a point with an image pixel value of 0 in the area, that is, to judge whether the image pixel values of the points in the area are valid pixel values, if there is a point with an image pixel value of 0 in the area, that is, the image pixel of the point in the area If the values are not valid pixel values, it is determined that there is occlusion between the upper body area and the lower body area.
在另一个实施场景中，可以通过将每个分割区域中具有相同像素值的像素值的点进行单像素轮廓提取获取每个分割区域中的轮廓的数量量。例如，获取每个分割区域中，具有相同像素值点构成的连通区域的轮廓的数量，若一个分割区域是无遮挡的，则其轮廓的数量量应该为1，若轮廓的数量量大于1，则表示该分割区域具有多个连通区域，即存在遮挡。In another implementation scenario, the number of contours in each segmented area may be obtained by performing single-pixel contour extraction on points with the same pixel value in each segmented area. For example, obtain the number of contours of connected regions with the same pixel value points in each segmented area. If a segmented area is unoccluded, the number of its contours should be 1. If the number of contours is greater than 1, It means that the segmentation region has multiple connected regions, that is, there is occlusion.
在一个实施场景中，可以结合有效像素值区域面积和/或区域连通性的判断结果来判断各个分割区域是否被遮挡，也可以仅仅有效像素值区域面积和区域连通性的的判断结果来判断各个分割区域是否被遮挡。In an implementation scenario, it can be combined with the judgment results of the effective pixel value area and/or regional connectivity to determine whether each segmented area is blocked, or only the judgment results of the effective pixel value area and regional connectivity can be used to judge whether each segmented area is blocked. Whether the segmented area is occluded.
在其他实施场景中，判定人体区域分割图像中至少一个分割区域存在遮挡，就可以判定人体图像存在遮挡，需要进一步获取人体图像的遮挡情况是否严重，若人体图像存在严重的遮挡情况，则需要对该人体图像进行处理，例如删除或者特殊标记等。可以通过在人体区域分割图像中的图像像素值为0的关键点位置的数量进行判断，若该数量超过预设数量阈值(例如，3个)，则表示人体有较大区域被遮挡了，判定人体图像的质量不满足预设要求。In other implementation scenarios, it can be determined that there is occlusion in at least one segmented area in the human body region segmentation image, and it can be determined that there is occlusion in the human body image. It is necessary to further obtain whether the occlusion of the human body image is serious. The human body image is processed, such as deletion or special marking. It can be judged by the number of key point positions with an image pixel value of 0 in the human body region segmentation image. If the number exceeds the preset number threshold (for example, 3), it means that a large area of the human body is blocked, and the judgment The quality of the human body image does not meet the preset requirements.
还可以通过获取每个分割区域的有效像素值区域面积，若存在至少一个分割区域的有效像素值面积为0，则表示该分割区完全被遮挡了，判定人体图像的质量不满足预设要求。It is also possible to obtain the effective pixel value area of each segmented area, if there is at least one segmented area with an effective pixel value area of 0, it means that the segmented area is completely blocked, and it is determined that the quality of the human body image does not meet the preset requirements.
还可以计算每个分割区域的轮廓的数量进行判断，若存在至少一个分割区域中的轮廓的数量大于3，则至少一个分割区域中存在贯穿、多块等情况的遮挡，判定人体图像的质量不满足预设要求。It is also possible to calculate the number of contours in each segmented region for judgment. If the number of contours in at least one segmented region is greater than 3, then at least one segmented region has occlusions such as penetration and multiple blocks, and it is determined that the quality of the human body image is not good. Meet preset requirements.
还可以断上半身区域和下半身区域之间的区域是否均为有效像素值，若上半身区域和下半身区域之间的区域不是均为有效像素值，则判定人体中间部分存在遮挡，且判定人体图像的质量不满足预设要求。It can also determine whether the area between the upper body area and the lower body area is a valid pixel value. If the area between the upper body area and the lower body area is not all valid pixel values, it is determined that the middle part of the human body is occluded, and the quality of the human body image is determined. Preset requirements are not met.
人体图像的质量不满足预设要求，则无法用于后续的人体识别等操作，可以删除该人体图像，重新获取新的人体图像进行上述步骤。If the quality of the human body image does not meet the preset requirements, it cannot be used for subsequent operations such as human body recognition. The human body image can be deleted, and a new human body image can be obtained again to perform the above steps.
通过上述描述可知，在本实施例中获取人体图像的关键点位置和语义分割结果，基于关键点位置和语义分割结果获取人体图像的人体区域分割图像，基于每个分割区域的有效像素值区域面积和/或区域连通性判断各个分割区域是否被遮挡，将关键点检测及图像分割方法结合进行遮挡检测，舍弃了关键点置信度而结合语义分割结果进行遮挡判断，不需要根据环境调整阈值，对环境鲁棒性高，检测结果准确且可靠。It can be seen from the above description that in this embodiment, the key point positions and semantic segmentation results of the human body image are obtained, the human body region segmentation images of the human body image are obtained based on the key point positions and the semantic segmentation results, and the effective pixel value region area of each segmented region is obtained And/or regional connectivity to judge whether each segmented area is occluded, the key point detection and image segmentation method are combined for occlusion detection, the key point confidence is discarded and the occlusion judgment is combined with the semantic segmentation results, and the threshold value does not need to be adjusted according to the environment. The environment is robust, and the detection results are accurate and reliable.
请参阅图6，图6是本发明提供的人体遮挡检测方法的第二实施例的流程示意图。本发明提供的人体遮挡检测方法包括如下步骤：Please refer to FIG. 6 . FIG. 6 is a schematic flowchart of a second embodiment of the method for detecting human occlusion provided by the present invention. The human body occlusion detection method provided by the present invention comprises the following steps:
S201：获取待处理图像，裁减出待处理图像中的人体图像。S201: Acquire an image to be processed, and cut out a human body image in the image to be processed.
S202：将人体图像输入神经网络模型，获取人体图像的关键点位置和语义分割结果，基于关键点位置和语义分割结果获取人体图像的人体区域分割图像，人体区域分割图像包括多个分割区域。S202: Input the human body image into the neural network model, obtain key point positions and semantic segmentation results of the human body image, and obtain human body region segmentation images of the human body image based on the key point positions and semantic segmentation results, where the human body region segmentation image includes multiple segmented regions.
S203：获取关键点位置在人体区域分割图像中的图像像素值，判断图像像素值是否为有效像素值；若是，执行步骤S204或S207，若否，执行步骤S208。S203: Obtain the image pixel value of the key point position in the human body region segmentation image, and judge whether the image pixel value is a valid pixel value; if yes, execute step S204 or S207, if not, execute step S208.
S204：判断每个分割区域的有效像素值区域面积是否低于对应的预设面积阈值。若否，执行步骤S205，若是，执行步骤S208。S204: Determine whether the effective pixel value region area of each segmented region is lower than the corresponding preset area threshold. If not, execute step S205, and if yes, execute step S208.
S205：将下半身区域和上半身区域的有效像素值区域面积相比，判断比值是否低于预设比值阈值，若否，执行步骤S206，若是，执行步骤S208。S205: Comparing the areas of effective pixel values of the lower body area and the upper body area, and determining whether the ratio is lower than a preset ratio threshold, if not, execute step S206, and if yes, execute step S208.
S206：判断上半身区域和下半身区域之间的区域是否均为有效像素值，若否，执行步骤S208。S206: Determine whether the area between the upper body area and the lower body area has valid pixel values, if not, go to step S208.
S207：获取每个分割区域中的轮廓的数量，判断分割区域中的轮廓的数量是否大于或等于2，若是，执行步骤S208。S207: Obtain the number of contours in each segmented region, and determine whether the number of contours in the segmented region is greater than or equal to 2, and if so, execute step S208.
S208：判定至少一个分割区域存在遮挡情况。S208: Determine that at least one segmented area is occluded.
在一个具体的实施场景中，步骤S201-S208的具体内容，已经在上文中进行具体阐述，此处不再进行赘述。In a specific implementation scenario, the specific content of steps S201-S208 has been described in detail above, and will not be repeated here.
请参阅图7，图7是本发明提供的人体遮挡检测方法的第三实施例的流程示意图。本发明提供的人体遮挡检测方法包括如下步骤：Please refer to FIG. 7 . FIG. 7 is a schematic flowchart of a third embodiment of a method for detecting human occlusion provided by the present invention. The human body occlusion detection method provided by the present invention comprises the following steps:
S301：获取待处理图像，裁减出待处理图像中的人体图像。S301: Acquire an image to be processed, and cut out a human body image in the image to be processed.
S302：将人体图像输入神经网络模型，获取人体图像的关键点位置和语义分割结果，基于关键点位置和语义分割结果获取人体图像的人体区域分割图像，人体区域分割图像包括多个分割区域，每个分割区域具有对应的像素值。S302: Input the human body image into the neural network model, obtain key point positions and semantic segmentation results of the human body image, and obtain human body region segmentation images of the human body image based on the key point positions and semantic segmentation results. The human body region segmentation image includes multiple segmented regions, each segmented regions have corresponding pixel values.
S303：获取关键点位置在人体区域分割图像中对应的点的图像像素值，判断图像像素值是否均为有效像素值；若是，执行步骤S305或S308，若否，执行步骤S304。S303: Obtain the image pixel value of the point corresponding to the key point position in the human body region segmentation image, and judge whether the image pixel values are all valid pixel values; if yes, execute step S305 or S308; if not, execute step S304.
S304：计算被遮挡的关键点位置的数量，判断数量是否超过预设数量阈值，若是，执行步骤S310。S304: Calculate the number of occluded key point positions, and determine whether the number exceeds a preset number threshold, and if so, execute step S310.
S305：判断每个分割区域的有效像素值区域面积是否低于对应的预设面积阈值。若是，执行步骤S306。若否，执行步骤S307。S305: Determine whether the effective pixel value region area of each segmented region is lower than the corresponding preset area threshold. If yes, execute step S306. If not, execute step S307.
S306：判断是否存在分割区域的有效像素值区域面积为0，若是，执行步骤S310。S306: Determine whether there is an effective pixel value area of the segmented area is 0, and if so, execute step S310.
S307：判断上半身区域和下半身区域之间的区域是否均为有效像素值，若否，执行步骤S310。S307: Determine whether the areas between the upper body area and the lower body area are valid pixel values, if not, go to step S310.
S308：获取每个分割区域中的连通区域的轮廓的数量的数量，判断分割区域中的轮廓的数量是否大于或等于2，若是，执行步骤S309。S308: Obtain the number of contours of connected regions in each segmented region, and determine whether the number of contours in the segmented region is greater than or equal to 2, and if so, perform step S309.
S309：判断分割区域中的轮廓的数量是否大于或等于3，若是，执行步骤S310。S309: Determine whether the number of contours in the segmented area is greater than or equal to 3, and if so, execute step S310.
S310：判定人体图像的质量不满足预设要求。S310: Determine that the quality of the human body image does not meet a preset requirement.
在一个具体的实施场景中，步骤S301-S310的具体内容，已经在上文中进行具体阐述，此处不再进行赘述。In a specific implementation scenario, the specific content of steps S301-S310 has been described in detail above, and will not be repeated here.
在一个实施场景中，将图1-图3中任一幅所示的方法在Partial-Reid数据集上进行测试，通过测试原图和可视化的结果图像进行对比能直观的判断对遮挡检测的准确性和有效性，与此同时还能准确的定位被遮挡的具体位置，测试结果如图8所示，图8是本发明提供的人体遮挡检测结果的示意图。In an implementation scenario, the method shown in any one of Figures 1-3 is tested on the Partial-Reid dataset, and the accuracy of occlusion detection can be intuitively judged by comparing the original test image with the visualized result image performance and effectiveness, and at the same time accurately locate the specific position that is blocked. The test results are shown in FIG. 8, which is a schematic diagram of the human body occlusion detection results provided by the present invention.
在一个实施场景中，在Partial-Reid数据集625张测试集上进行了测试，其中遮挡测试集467张，能检测出遮挡图像451张；非遮挡测试集158张，能检测出非遮挡158张，其测试结果如下表1：In an implementation scenario, tests were carried out on 625 test sets of the Partial-Reid dataset, including 467 occlusion test sets, 451 occlusion images can be detected; 158 non-occlusion test sets, 158 non-occlusion images can be detected , the test results are shown in Table 1:
表1测试结果Table 1 Test results
由表1可以看出，以遮挡行人总共467，能准确检测出463个，其检测正确率达到99.1％，非遮挡正常行人总共158个，能正确检测出非遮挡数量158个，正确率达到100％。It can be seen from Table 1 that 467 occluded pedestrians can be detected accurately, and the detection accuracy rate reaches 99.1%. There are 158 non-occluded normal pedestrians, and 158 non-occluded pedestrians can be detected correctly, and the correct rate reaches 100%. %.
本发明不仅解决了传统及基于分类算法对遮挡检测精度不高的问题，还能以结果可视化解释遮挡检测，提高对人体遮挡检测方法的可用性，且相对于关键点置信度方法，本申请通过将关键点检测和图像分割技术相结合，不需要根据环境调整阈值，应用简单，对环境鲁棒性高；测过滤遮挡行人，在很大程度上提升如reid(Re-identification，利用算法)行人识别、行人跟踪、工服识别等方法的准确性。The invention not only solves the problem of low occlusion detection accuracy of traditional and classification-based algorithms, but also can explain the occlusion detection with the result visualization, which improves the usability of the human body occlusion detection method, and compared with the key point confidence method, this application adopts the The combination of key point detection and image segmentation technology does not need to adjust the threshold according to the environment, the application is simple, and it is highly robust to the environment; measuring and filtering to block pedestrians can greatly improve pedestrian recognition such as reid (Re-identification, using algorithms) , Pedestrian tracking, work clothes recognition and other methods of accuracy.
请参阅图9，图9是本发明提供的计算机设备的一实施例的结构示意图。智能设备10包括处理器11、存储器12。处理器11耦接存储器12。存储器12中存储有计算机程序，处理器11在工作时执行该计算机程序以实现如上的方法。详细的步骤可参见上述，在此不再赘述。Please refer to FIG. 9 . FIG. 9 is a schematic structural diagram of an embodiment of a computer device provided by the present invention. The
请参阅图10，图10是本发明提供的计算机可读存储介质的一实施例的结构示意图。计算机可读存储介质20中存储有至少一个计算机程序21，计算机程序21用于被处理器执行以实现如上方法，详细的步骤可参见上述，在此不再赘述。在一个实施例中，计算机可读存储介质20可以是终端中的存储芯片、硬盘或者是移动硬盘或者优盘、光盘等其他可读写存储的工具，还可以是服务器等等。Please refer to FIG. 10 . FIG. 10 is a schematic structural diagram of an embodiment of a computer-readable storage medium provided by the present invention. At least one
所述存储介质可以由任何类型的易失性或非易失性存储设备、或者它们的组合来实现。其中，非易失性存储器可以是只读存储器(ROM，Read Only Memory)、可编程只读存储器(PROM，Programmable Read-Only Memory)、可擦除可编程只读存储器(EPROM，ErasableProgrammable Read-Only Memory)、电可擦除可编程只读存储器(EEPROMElectricallyErasable Programmable Read-Only Memory)、磁性随机存取存储器(FRAM，FerromagneticRandom Access Memory)、快闪存储器(Flash Memory)、磁表面存储器、光盘、或只读光盘(CD-ROM，Compact Disc Read-Only Memory)；磁表面存储器可以是磁盘存储器或磁带存储器。易失性存储器可以是随机存取存储器(RAM，Random AccessMemory)，其用作外部高速缓存。通过示例性但不是限制性说明，许多形式的RAM可用，例如静态随机存取存储器(SRAM，Static Random Access Memory)、同步静态随机存取存储器(SSRAM，SynchronousStatic Random Access Memory)、动态随机存取存储器(DRAM，Dynamic Random AccessMemory)、同步动态随机存取存储器(SDRAM，Synchronous DynamicRandomAccessMemory)、双倍数据速率同步动态随机存取存储器(DDRSDRAM，Double DataRateSynchronous Dynamic Random Access Memory)、增强型同步动态随机存取存储器(ESDRAM，Enhanced Synchronous Dynamic Random Access Memory)、同步连接动态随机取存储器(SLDRAM，SyncLink Dynamic Random Access Memory)、直接内存总线随机存取存储器(DRRAM，Direct Rambus Random Access Memory)。本发明实施例描述的存储介质旨在包括但不限于这些和任意其它适合类型的存储器。The storage medium may be implemented by any type of volatile or non-volatile storage device, or a combination thereof. Wherein, non-volatile memory can be read-only memory (ROM, Read Only Memory), programmable read-only memory (PROM, Programmable Read-Only Memory), erasable programmable read-only memory (EPROM, ErasableProgrammable Read-Only Memory) Memory), Electrically Erasable Programmable Read-Only Memory (EEPROM ElectricallyErasable Programmable Read-Only Memory), Magnetic Random Access Memory (FRAM, Ferromagnetic Random Access Memory), Flash Memory (Flash Memory), Magnetic Surface Memory, CD-ROM, or only Read CD (CD-ROM, Compact Disc Read-Only Memory); magnetic surface storage can be disk storage or tape storage. The volatile memory may be random access memory (RAM, Random Access Memory), which is used as an external cache. By way of illustration and not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM, Static Random Access Memory), Synchronous Static Random Access Memory (SSRAM, Synchronous Static Random Access Memory), Dynamic Random Access Memory (DRAM, Dynamic Random Access Memory), Synchronous Dynamic Random Access Memory (SDRAM, Synchronous Dynamic Random Access Memory), Double Data Rate Synchronous Dynamic Random Access Memory (DDRSDRAM, Double DataRateSynchronous Dynamic Random Access Memory), Enhanced Synchronous Dynamic Random Access Memory (ESDRAM, Enhanced Synchronous Dynamic Random Access Memory), Synchronous Link Dynamic Random Access Memory (SLDRAM, SyncLink Dynamic Random Access Memory), Direct Memory Bus Random Access Memory (DRRAM, Direct Rambus Random Access Memory). The storage media described in the embodiments of the present invention are intended to include, but are not limited to, these and any other suitable types of memory.
在本发明所提供的几个实施例中，应该理解到，所揭露的系统和方法，可以通过其它的方式实现。以上所描述的设备实施例仅仅是示意性的，例如，所述单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，如：多个单元或组件可以结合，或可以集成到另一个系统，或一些特征可以忽略，或不执行。另外，所显示或讨论的各组成部分相互之间的耦合、或直接耦合、或通信连接可以是通过一些接口，设备或单元的间接耦合或通信连接，可以是电性的、机械的或其它形式的。In the several embodiments provided by the present invention, it should be understood that the disclosed system and method can be implemented in other ways. The device embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods, such as: multiple units or components can be combined, or May be integrated into another system, or some features may be ignored, or not implemented. In addition, the coupling, or direct coupling, or communication connection between the components shown or discussed may be through some interfaces, and the indirect coupling or communication connection of devices or units may be electrical, mechanical or other forms of.
上述作为分离部件说明的单元可以是、或也可以不是物理上分开的，作为单元显示的部件可以是、或也可以不是物理单元，即可以位于一个地方，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全部单元来实现本实施例方案的目的。The units described above 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 distributed to multiple network units; Part or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.
另外，在本发明各实施例中的各功能单元可以全部集成在一个处理单元中，也可以是各单元分别单独作为一个单元，也可以两个或两个以上单元集成在一个单元中；上述集成的单元既可以采用硬件的形式实现，也可以采用硬件加软件功能单元的形式实现。In addition, each functional unit in each embodiment of the present invention can be integrated into one processing unit, or each unit can be used as a single unit, or two or more units can be integrated into one unit; the above-mentioned integration The unit can be realized in the form of hardware or in the form of hardware plus software functional unit.
本领域普通技术人员可以理解：实现上述方法实施例的全部或部分步骤可以通过程序指令相关的硬件来完成，前述的程序可以存储于一计算机可读取存储介质中，该程序在执行时，执行包括上述方法实施例的步骤；而前述的存储介质包括：移动存储设备、只读存储器(ROM，Read-Only Memory)、随机存取存储器(RAM，Random Access Memory)、磁碟者光盘等各种可以存储程序代码的介质。Those of ordinary skill in the art can understand that all or part of the steps to realize the above method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, the Include the steps of the foregoing method embodiments; and the aforementioned storage medium includes: various A medium on which program code can be stored.
或者，本发明上述集成的单元如果以软件功能模块的形式实现并作为独立的产品销售或使用时，也可以存储在一个计算机可读取存储介质中。基于这样的理解，本发明实施例的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可以是个人计算机、服务器、或者网络设备等)执行本发明各个实施例所述方法的全部或部分。而前述的存储介质包括：移动存储设备、ROM、RAM、磁碟或者光盘等各种可以存储程序代码的介质。Alternatively, if the above-mentioned integrated units of the present invention are implemented in the form of software function modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for Make a computer device (which may be a personal computer, a server, or a network device, etc.) execute all or part of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media capable of storing program codes such as removable storage devices, ROM, RAM, magnetic disks or optical disks.
本发明所提供的几个方法实施例中所揭露的方法，在不冲突的情况下可以任意组合，得到新的方法实施例。The methods disclosed in the several method embodiments provided by the present invention can be combined arbitrarily under the condition of no conflict to obtain new method embodiments.
本发明所提供的几个产品实施例中所揭露的特征，在不冲突的情况下可以任意组合，得到新的产品实施例。The features disclosed in several product embodiments provided by the present invention can be combined arbitrarily without conflict to obtain new product embodiments.
本发明所提供的几个方法或设备实施例中所揭露的特征，在不冲突的情况下可以任意组合，得到新的方法实施例或设备实施例。The features disclosed in several method or device embodiments provided by the present invention can be combined arbitrarily without conflict to obtain new method embodiments or device embodiments.
以上内容是结合具体的优选实施方式对本发明所做的进一步详细说明，不能认定本发明的具体实施只局限于这些说明。对于本发明所属技术领域的技术人员来说，在不脱离本发明构思的前提下，还可以做出若干等同替代或明显变型，而且性能或用途相同，都应当视为属于本发明的保护范围。The above content is a further detailed description of the present invention in conjunction with specific preferred embodiments, and it cannot be assumed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art to which the present invention belongs, several equivalent substitutions or obvious modifications can be made without departing from the concept of the present invention, and those with the same performance or use should be deemed to belong to the protection scope of the present invention.
Claims (10)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310090877.3ACN116363692A (en) | 2023-01-18 | 2023-01-18 | Human body shielding detection method, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310090877.3ACN116363692A (en) | 2023-01-18 | 2023-01-18 | Human body shielding detection method, computer equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116363692Atrue CN116363692A (en) | 2023-06-30 |
Family
ID=86905393
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310090877.3APendingCN116363692A (en) | 2023-01-18 | 2023-01-18 | Human body shielding detection method, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116363692A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN119672764A (en)* | 2023-12-15 | 2025-03-21 | 徐州宏协文化传媒有限公司 | Intelligent online video teaching effect management system |
| CN119693403A (en)* | 2025-02-21 | 2025-03-25 | 光轮智能(北京)科技有限公司 | Contour extraction method, computer device and medium for semantic segmentation map |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109766868A (en)* | 2019-01-23 | 2019-05-17 | 哈尔滨工业大学 | A real scene occluded pedestrian detection network and detection method based on body key point detection |
| CN109902593A (en)* | 2019-01-30 | 2019-06-18 | 济南大学 | A Kinect-based Gesture Occlusion Detection Method and System |
| CN111310624A (en)* | 2020-02-05 | 2020-06-19 | 腾讯科技（深圳）有限公司 | Occlusion recognition method and device, computer equipment and storage medium |
| CN111753724A (en)* | 2020-06-24 | 2020-10-09 | 上海依图网络科技有限公司 | Abnormal behavior identification method and device |
| CN112907702A (en)* | 2020-12-07 | 2021-06-04 | 腾讯科技（深圳）有限公司 | Image processing method, image processing device, computer equipment and storage medium |
| CN112927262A (en)* | 2021-03-22 | 2021-06-08 | 瓴盛科技有限公司 | Camera lens shielding detection method and system based on video |
| CN113706506A (en)* | 2021-08-27 | 2021-11-26 | 创新奇智(重庆)科技有限公司 | Method and device for detecting assembly state, electronic equipment and storage medium |
- 2023
- 2023-01-18CNCN202310090877.3Apatent/CN116363692A/enactivePending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109766868A (en)* | 2019-01-23 | 2019-05-17 | 哈尔滨工业大学 | A real scene occluded pedestrian detection network and detection method based on body key point detection |
| CN109902593A (en)* | 2019-01-30 | 2019-06-18 | 济南大学 | A Kinect-based Gesture Occlusion Detection Method and System |
| CN111310624A (en)* | 2020-02-05 | 2020-06-19 | 腾讯科技（深圳）有限公司 | Occlusion recognition method and device, computer equipment and storage medium |
| CN111753724A (en)* | 2020-06-24 | 2020-10-09 | 上海依图网络科技有限公司 | Abnormal behavior identification method and device |
| CN112907702A (en)* | 2020-12-07 | 2021-06-04 | 腾讯科技（深圳）有限公司 | Image processing method, image processing device, computer equipment and storage medium |
| CN112927262A (en)* | 2021-03-22 | 2021-06-08 | 瓴盛科技有限公司 | Camera lens shielding detection method and system based on video |
| CN113706506A (en)* | 2021-08-27 | 2021-11-26 | 创新奇智(重庆)科技有限公司 | Method and device for detecting assembly state, electronic equipment and storage medium |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN119672764A (en)* | 2023-12-15 | 2025-03-21 | 徐州宏协文化传媒有限公司 | Intelligent online video teaching effect management system |
| CN119693403A (en)* | 2025-02-21 | 2025-03-25 | 光轮智能(北京)科技有限公司 | Contour extraction method, computer device and medium for semantic segmentation map |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109670441B (en) | Method, system, terminal and computer readable storage medium for realizing wearing recognition of safety helmet | |
| CN107358149B (en) | Human body posture detection method and device | |
| CN116363692A (en) | Human body shielding detection method, computer equipment and storage medium | |
| JP7235892B2 (en) | Image processing method, apparatus, electronic equipment and computer program product | |
| US20150286281A1 (en) | Generating a screenshot | |
| CN109918969A (en) | Face detection method and device, computer device and computer-readable storage medium | |
| CN110610127A (en) | Face recognition method and device, storage medium and electronic equipment | |
| CN105069431B (en) | The localization method and device of face | |
| CN112464797B (en) | Smoking behavior detection method and device, storage medium and electronic equipment | |
| CN110781770B (en) | Living body detection method, device and equipment based on face recognition | |
| CN110222621B (en) | Similar house type detection method, electronic equipment and storage medium | |
| CN112348112B (en) | Training method and training device for image recognition model and terminal equipment | |
| CN116188379A (en) | Edge defect detection method, device, electronic equipment and storage medium | |
| CN110415237A (en) | Skin flaw detection method, skin flaw detection device, terminal device and readable storage medium | |
| CN113392681A (en) | Human body falling detection method and device and terminal equipment | |
| CN109375833B (en) | Touch instruction generation method and device | |
| CN114387306A (en) | Target displacement monitoring method and device, computer equipment and storage medium | |
| WO2021012513A1 (en) | Gesture operation method and apparatus, and computer device | |
| CN111626265A (en) | Multi-camera downlink identification method and device and computer readable storage medium | |
| CN113743195B (en) | Face occlusion quantitative analysis method, device, electronic device and storage medium | |
| CN118567486B (en) | Digital sand table interaction method and system | |
| CN112183277A (en) | Detection method and device for abandoned object and lost object, terminal equipment and storage medium | |
| CN111626299A (en) | Outline-based digital character recognition method | |
| CN116912945A (en) | Gesture recognition method, device, equipment and computer program product | |
| CN111080664B (en) | Data processing method and device, computer storage medium and computer |
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 |