WO2018121127A1 - System for collecting statistics on pedestrian traffic by means of tracking based on video analysis technique - Google Patents

System for collecting statistics on pedestrian traffic by means of tracking based on video analysis technique Download PDF

Info

Publication number
WO2018121127A1
WO2018121127A1 PCT/CN2017/111929 CN2017111929W WO2018121127A1 WO 2018121127 A1 WO2018121127 A1 WO 2018121127A1 CN 2017111929 W CN2017111929 W CN 2017111929W WO 2018121127 A1 WO2018121127 A1 WO 2018121127A1
Authority
WO
WIPO (PCT)
Prior art keywords
module
human body
trajectory
contour
recognition model
Prior art date
Application number
PCT/CN2017/111929
Other languages
French (fr)
Chinese (zh)
Inventor
周圣强
Original Assignee
苏州万店掌网络科技有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 苏州万店掌网络科技有限公司 filed Critical 苏州万店掌网络科技有限公司
Publication of WO2018121127A1 publication Critical patent/WO2018121127A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/50—Context or environment of the image
    • G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
    • G06V20/53—Recognition of crowd images, e.g. recognition of crowd congestion

Definitions

  • the present invention relates to the field of video analysis technologies, and in particular, to a system for tracking statistic traffic based on video analysis technology.
  • the traditional human traffic statistics are two kinds of infrared counters and mechanical counters.
  • the infrared counters use infrared sensors to sense the statistics of the personnel, but when there are many people passing by, some people are difficult to be sensed by the infrared sensor because of the occlusion. And the infrared counter intelligently recognizes that someone enters and exits, cannot determine the direction, is prone to false positives, and has low accuracy.
  • the mechanical counter needs to embed the induction pedal in the passage port, and the counter is used to count the personnel counting. This method is inaccurate when the flow rate is large, and it is only a rough statistical number of people passing through. The person is entering or going out.
  • the feature points of some motions are tracked first, then the trajectories of the feature points are clustered and analyzed, so that the flow information is obtained.
  • the feature points themselves are difficult to track stably, and the technical precision is poor.
  • the second is based on the human body segmentation tracking method, first extracting the moving target block, then segmenting the moving target block to obtain a single human target, and finally tracking each human target to achieve human flow statistics. If the human body is in the occlusion state, the accuracy of human body segmentation is difficult to be guaranteed, which affects statistical accuracy.
  • the third method is based on the detection and tracking of the head or the head and shoulder.
  • the method detects the head or the head and shoulders in the video, and performs the flow statistics by tracking the head or the head and shoulders. This method can only detect the same type of target, and cannot detect different types of targets at the same time.
  • the object of the present invention is to overcome the above problems existing in the prior art, and to provide a system for tracking statistic person traffic based on a video analysis mode.
  • the system of the present invention can not only accurately identify people entering and exiting, but also can determine direction and false positive rate. Lower.
  • a system for tracking statistic traffic based on video analytics technology comprising:
  • the storage module stores a pre-established human body recognition model, and the storage module further stores a motion track preset in advance;
  • the comparison module compares image information acquired by the camera with a human body recognition model stored in the storage module;
  • a positioning module after the comparison module performs the comparison of the human body recognition module, if the acquired image information is successfully compared with the human body recognition model, the first step is performed;
  • trajectory forming module tracks the recognized human body image information based on the first step of the positioning module to form a human body moving trajectory in a three-dimensional space
  • the direction module compares a human body moving track formed by the track forming module with a preset motion track in the storage module, and determines a direction of a human body moving track;
  • the counting module counts the flow of the human body according to the direction of the human body moving track determined by the direction module;
  • the calculation module performs statistical calculation on the incoming traffic and the output traffic calculated by the counting module, and then analyzes with the background system data, and performs calculation according to the sales amount and the sales order number to obtain the conversion rate of the human flow.
  • the method further includes an extraction module, the extraction module extracts a human body contour from the image information acquired by the camera, and transmits the contour to the comparison module, where the comparison module extracts the human body contour extracted by the extraction module Compare with the stored human body recognition model.
  • the comparison module further comprises, when comparing the extracted human body contour with the stored human body recognition module, a threshold analysis of the human body contour contrast similarity, wherein the similarity between the human body contour and the stored human body recognition model is higher than a threshold value.
  • a threshold analysis of the human body contour contrast similarity wherein the similarity between the human body contour and the stored human body recognition model is higher than a threshold value.
  • the trajectory forming module defines an entering direction as X, and an outgoing direction as Y, and after the first positioning by the positioning module, respectively reduces or increases the trajectory in two directions of X and Y. , to calculate the movement trajectory.
  • the trajectory forming module further performs smoothness analysis on the human body moving trajectory, determines whether the smoothness satisfies the threshold of the pre-custom trajectory, and if so, retains the human body moving trajectory, and if not, quits the human body moving trajectory.
  • the scene dividing module further performs scene division on the detection area in the image, and obtains a variation range of the size of the human body image in the detection area by scene division.
  • the calculation module forms a report for the statistics of the flow of the person, and analyzes the operation status of the store according to the report.
  • the system of the invention first establishes a human body model in advance, and then compares the comparison module, first determines the acquired image as a person, reduces the false positive rate of the system; first customizes a preset trajectory, obtains a moving trajectory in a three-dimensional space, and compares the trajectory according to the trajectory. To determine the direction of people entering and exiting, it is possible to accurately identify the entry and exit of people, and also to determine the direction with high precision.
  • the system of the invention can automatically generate reports, obtain data such as passenger flow, customer volume, number of stops, etc., and can comprehensively analyze data such as sales and sales orders in the background, obtain data information such as passenger flow conversion rate, and analyze the merchants.
  • the store situation is of great significance.
  • Figure 1 is a schematic diagram of the system of the present invention.
  • the hardware mainly includes a camera, a storage device, a CPU computing device, and the like.
  • the functions include: a storage module, an extraction module, a comparison module, a positioning module, a track formation module, a direction module, a counting module, a calculation module, and a scene division module.
  • the camera is used to obtain image information, and the camera setting position is set according to a specific scene and stored in the storage module.
  • a pre-established human body recognition model and storing a motion trajectory of the pre-customized setting in the storage module;
  • the comparison module compares the image information acquired by the camera with the human body recognition model stored in the storage module; before the comparison, the human body contour is first performed
  • the extraction and extraction module extracts the human body contour from the image information acquired by the camera, and transmits the contour to the comparison module, and the comparison module compares the human body contour extracted by the extraction module with the stored human body recognition model.
  • the contrast module compares the extracted human body contour with the stored human body recognition module, and also includes a threshold analysis of the human body contour contrast similarity.
  • the human body contour and the stored human body recognition model have a similarity similar to the threshold value
  • the human body contour and the human body contour When the recognition model is successfully matched, when the similarity between the human body contour and the stored human body recognition module is lower than the threshold value, the human body contour fails to match the human body recognition model, and the image acquired by the camera cannot be defined as a person.
  • the trajectory forming module tracks the human body image information after the recognition based on the first step of the positioning module to form The movement of the human body in three-dimensional space.
  • the trajectory forming module defines the entering direction as X and the outgoing direction as Y. After the first positioning by the positioning module, the movement is calculated by reducing or increasing the trajectory in the X and Y directions respectively. Track.
  • the trajectory forming module also performs smoothness analysis on the trajectory of the human body to determine whether the smoothness satisfies the threshold of the pre-custom trajectory. If it is satisfied, the trajectory of the human body is retained, and if not, the trajectory of the human body is discarded.
  • the direction module compares the movement track of the human body formed by the track forming module with the preset motion track in the storage module to determine the direction of the human body moving track; then the counting module counts the human flow according to the direction of the human body moving track determined by the direction module; The module counts the incoming traffic and the output traffic for statistical calculation, and then analyzes with the background system data, and calculates according to the sales amount and the sales order number to obtain the conversion rate of the person flow.
  • the calculation module forms a report for the statistics of the flow of people, and analyzes the operation status of the store according to the report.
  • the scene dividing module performs scene division on the detection area in the image, and obtains a variation range of the size of the human body image in the detection area by scene division.
  • the camera acquires the human body image in the scene division module, and then extracts the human body contour
  • the extraction module extracts the human body contour from the image information acquired by the camera, and transmits
  • the comparison module compares the human body contour extracted by the extraction module with the stored human body recognition model.
  • the comparison module performs the human body recognition module comparison, if the acquired image information and the human body recognition model Contrast success, first step positioning; track formation The module is based on the first step of the positioning module to track the recognized human body image information to form a human body moving trajectory in a three-dimensional space; perform smoothness analysis on the human body moving trajectory to determine whether the smoothness satisfies the threshold of the pre-customized trajectory, If satisfied, the human body movement track is retained, and if not satisfied, the human body moving track is discarded; the direction module compares the human body moving track formed by the track forming module with the preset motion track in the storage module to determine the direction

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Image Analysis (AREA)

Abstract

Disclosed is a system for collecting statistics on pedestrian traffic by means of tracking based on a video analysis technique. The system comprises: a camera, a storage module, a comparison module, a positioning module, a trajectory formation module, a directional module, a counting module, a calculation module, a pre-built body recognition model and a motion trajectory, wherein the comparison module compares image information obtained by the camera with a stored body recognition model, and if comparison is successful, first-step positioning is carried out; the trajectory formation module tracks, based on the first-step positioning, recognised body image information to form a body motion trajectory in a three-dimensional space; the directional module compares the body motion trajectory with a pre-set motion trajectory in the storage module to determine the direction of the body motion trajectory; and the counting module collects statistics on pedestrian traffic according to the direction of the body motion trajectory. The system in the present invention can not only accurately recognise that there are people coming in and going out, but also determine the direction, so a false positive rate is relatively low.

Description

一种基于视频分析技术的跟踪统计人流量的系统A system for tracking statistical traffic based on video analysis technology 技术领域Technical field
本发明涉及视频分析技术领域，具体涉及一种基于视频分析技术的跟踪统计人流量的系统。The present invention relates to the field of video analysis technologies, and in particular, to a system for tracking statistic traffic based on video analysis technology.
背景技术Background technique
随着视频监控技术的发展，在很多公共场所的出入口都设置了视频监控系统，以便管理人员对这些公共场所出入口进行监控和统计人流量。在超市、店铺、商场等消费场所，人流量对管理者有重要的意义。With the development of video surveillance technology, video surveillance systems have been set up at the entrances and exits of many public places, so that managers can monitor and count the traffic of these public places. In supermarkets, shops, shopping malls and other consumer places, the flow of people has important implications for managers.
传统的人流量统计是采用红外计数器和机械计数器两种，红外计数器利用红外感应器感应经过人员进行计数统计，但是当通过人员比较多时，有些人员因为遮挡，很难被红外感应器感测到，并且红外计数器智能辨识有人进出，无法确定方向，容易误报，精准度较低。机械式计数器需要在通道口埋设感应踏板，通过踩踏板触动计数器进行人员计数统计，这种方式在人流量较大的时候，计数不准确，而且也只是粗略的统计经过的人数，而无法确定经过的人员是进入还是出去。The traditional human traffic statistics are two kinds of infrared counters and mechanical counters. The infrared counters use infrared sensors to sense the statistics of the personnel, but when there are many people passing by, some people are difficult to be sensed by the infrared sensor because of the occlusion. And the infrared counter intelligently recognizes that someone enters and exits, cannot determine the direction, is prone to false positives, and has low accuracy. The mechanical counter needs to embed the induction pedal in the passage port, and the counter is used to count the personnel counting. This method is inaccurate when the flow rate is large, and it is only a rough statistical number of people passing through. The person is entering or going out.
为了解决上述问题，使用了基于视频分析的人流量统计方法，目前，基于视频分析的流量统计方法主要有三类：In order to solve the above problems, a human traffic statistics method based on video analysis is used. Currently, there are three main types of traffic statistics methods based on video analysis:
一是基于特征点跟踪的方法，先跟踪一些运动的特征点，然后对特征点的轨迹进行聚类分析，从而得到人流量信息，特征点本身难以稳定地跟踪，技术精度较差。Firstly, based on the method of feature point tracking, the feature points of some motions are tracked first, then the trajectories of the feature points are clustered and analyzed, so that the flow information is obtained. The feature points themselves are difficult to track stably, and the technical precision is poor.
二是基于人体分割的跟踪方法，首先提取出运动目标块，然后对运动目标块进行分割得到单个人体目标，最后跟踪各个人体目标实现人流量统计。如果当人体处于遮挡状态时，人体分割的准确性难以得到保证，影响统计精度。The second is based on the human body segmentation tracking method, first extracting the moving target block, then segmenting the moving target block to obtain a single human target, and finally tracking each human target to achieve human flow statistics. If the human body is in the occlusion state, the accuracy of human body segmentation is difficult to be guaranteed, which affects statistical accuracy.
三是基于人头或者头肩检测和跟踪的方法，该方法在视频中检测人头或者头肩，通过对人头或者头肩的跟踪进行人流量统计。该方法只能检测同一类目标，无法同时检测不同类目标。The third method is based on the detection and tracking of the head or the head and shoulder. The method detects the head or the head and shoulders in the video, and performs the flow statistics by tracking the head or the head and shoulders. This method can only detect the same type of target, and cannot detect different types of targets at the same time.
发明内容Summary of the invention
本发明的目的在于克服现有技术存在的以上问题，提供一种基于视频分析模式的跟踪统计人流量的系统，本发明的系统不但能够准确的辨识有人进出，而且还能够确定方向，误报率较低。 The object of the present invention is to overcome the above problems existing in the prior art, and to provide a system for tracking statistic person traffic based on a video analysis mode. The system of the present invention can not only accurately identify people entering and exiting, but also can determine direction and false positive rate. Lower.
为实现上述技术目的，达到上述技术效果，本发明通过以下技术方案实现：In order to achieve the above technical effects and achieve the above technical effects, the present invention is achieved by the following technical solutions:
一种基于视频分析技术的跟踪统计人流量的系统，其包括：A system for tracking statistic traffic based on video analytics technology, comprising:
摄像头，所述摄像头获取图像信息；a camera that acquires image information;
存储模块，所述存储模块存储预先建立的人体识别模型，所述存储模块还存储预先自定义设置的运动轨迹；a storage module, the storage module stores a pre-established human body recognition model, and the storage module further stores a motion track preset in advance;
对比模块，所述对比模块将所述摄像头获取的图像信息与所述存储模块中存储的人体识别模型进行对比；a comparison module, the comparison module compares image information acquired by the camera with a human body recognition model stored in the storage module;
定位模块，所述对比模块进行人体识别模块对比后，如果获取的图像信息与人体识别模型对比成功，进行第一步定位；a positioning module, after the comparison module performs the comparison of the human body recognition module, if the acquired image information is successfully compared with the human body recognition model, the first step is performed;
轨迹形成模块，所述轨迹形成模块基于所述定位模块的第一步定位，对识别后的人体图像信息进行跟踪，形成三维空间中的人体移动轨迹；a trajectory forming module, wherein the trajectory forming module tracks the recognized human body image information based on the first step of the positioning module to form a human body moving trajectory in a three-dimensional space;
方向模块，所述方向模块将所述轨迹形成模块形成的人体移动轨迹与所述存储模块中预先设置的运动轨迹做对比，确定人体移动轨迹的方向；a direction module, the direction module compares a human body moving track formed by the track forming module with a preset motion track in the storage module, and determines a direction of a human body moving track;
计数模块，所述计数模块根据所述方向模块确定的人体移动轨迹方向统计人流量；a counting module, wherein the counting module counts the flow of the human body according to the direction of the human body moving track determined by the direction module;
计算模块，所述计算模块对所述计数模块统计的进入人流量和输出人流量进行统计计算，然后与后台系统数据进行分析，根据销售额和销售单数进行计算，得到人流转化率。a calculation module, the calculation module performs statistical calculation on the incoming traffic and the output traffic calculated by the counting module, and then analyzes with the background system data, and performs calculation according to the sales amount and the sales order number to obtain the conversion rate of the human flow.
进一步优选地，还包括提取模块，所述提取模块从所述摄像头获取的图像信息中，将人体轮廓提取出来，并且传输至所述对比模块，所述对比模块将所述提取模块提取的人体轮廓与存储的人体识别模型进行对比。Further preferably, the method further includes an extraction module, the extraction module extracts a human body contour from the image information acquired by the camera, and transmits the contour to the comparison module, where the comparison module extracts the human body contour extracted by the extraction module Compare with the stored human body recognition model.
进一步优选地，所述对比模块在将提取的人体轮廓与存储的人体识别模块进行对比时，还包括对人体轮廓对比相似度阈值分析，当人体轮廓与存储的人体识别模型对比相似度高于阈值时，人体轮廓与人体识别模型匹配成功，当人体轮廓与存储的人体识别模块对比相似度低于阈值时，人体轮廓与人体识别模型匹配失败，不能将所述摄像头获取的图像定义为人。Further preferably, the comparison module further comprises, when comparing the extracted human body contour with the stored human body recognition module, a threshold analysis of the human body contour contrast similarity, wherein the similarity between the human body contour and the stored human body recognition model is higher than a threshold value. When the human body contour is successfully matched with the human body recognition model, when the similarity between the human body contour and the stored human body recognition module is lower than the threshold value, the human body contour fails to match the human body recognition model, and the image acquired by the camera cannot be defined as a person.
进一步优选地，所述轨迹形成模块将进入方向定义为X，将出去方向定义为Y，通过所述定位模块进行第一步定位后，分别通过在X、Y两个方向上轨迹的减少或增加，来计算得到移动轨迹。Further preferably, the trajectory forming module defines an entering direction as X, and an outgoing direction as Y, and after the first positioning by the positioning module, respectively reduces or increases the trajectory in two directions of X and Y. , to calculate the movement trajectory.
进一步优选地，所述轨迹形成模块还对人体移动轨迹进行平滑度分析，判断平滑度是否满足预先自定义轨迹的阈值，如果满足，保留人体移动轨迹，如果不满足，则放弃人体移动轨迹。 Further preferably, the trajectory forming module further performs smoothness analysis on the human body moving trajectory, determines whether the smoothness satisfies the threshold of the pre-custom trajectory, and if so, retains the human body moving trajectory, and if not, quits the human body moving trajectory.
进一步优选地，还包括场景划分模块，所述场景划分模块对图像中的检测区域进行场景划分，通过场景划分，获取检测区域内人体图像尺寸的变化范围。Further preferably, the scene dividing module further performs scene division on the detection area in the image, and obtains a variation range of the size of the human body image in the detection area by scene division.
进一步优选地，所述计算模块针对人流量统计情况形成报表，根据报表后续分析店铺经营状况。Further preferably, the calculation module forms a report for the statistics of the flow of the person, and analyzes the operation status of the store according to the report.
本发明的有益效果是:The beneficial effects of the invention are:
本发明的系统预先先建立人体模型，然后通过对比模块进行对比，先确定获取的图像为人，降低系统的误报率；先自定义预设一个轨迹，在三维空间中获得移动轨迹，根据轨迹对比，确定人进出方向，能够准确的辨识有人进出，而且还能够确定方向，精度较高。The system of the invention first establishes a human body model in advance, and then compares the comparison module, first determines the acquired image as a person, reduces the false positive rate of the system; first customizes a preset trajectory, obtains a moving trajectory in a three-dimensional space, and compares the trajectory according to the trajectory. To determine the direction of people entering and exiting, it is possible to accurately identify the entry and exit of people, and also to determine the direction with high precision.
本发明的系统能够自动生成报表，获取进客流、出客量、驻足次数等数据，并且可以与后台的销售额、销售单数等数据进行综合分析，获得客流量转化率等数据信息，对商家分析店铺情况具有重要意义。The system of the invention can automatically generate reports, obtain data such as passenger flow, customer volume, number of stops, etc., and can comprehensively analyze data such as sales and sales orders in the background, obtain data information such as passenger flow conversion rate, and analyze the merchants. The store situation is of great significance.
上述说明仅是本发明技术方案的概述，为了能够更清楚了解本发明的技术手段，并可依照说明书的内容予以实施，以下以本发明的较佳实施例并配合附图详细说明如后。本发明的具体实施方式由以下实施例及其附图详细给出。The above description is only an overview of the technical solutions of the present invention, and the technical means of the present invention can be more clearly understood and can be implemented in accordance with the contents of the specification. Hereinafter, the preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. Specific embodiments of the present invention are given in detail by the following examples and the accompanying drawings.
附图说明DRAWINGS
为了更清楚地说明本发明实施例技术中的技术方案，下面将对实施例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, the drawings used in the technical description of the embodiments will be briefly described below. It is obvious that the drawings in the following description are only some implementations of the present invention. For example, other drawings may be obtained from those of ordinary skill in the art in light of the inventive work.
图1是本发明系统原理图。Figure 1 is a schematic diagram of the system of the present invention.
具体实施方式detailed description
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 technical solutions in the embodiments of the present invention are clearly and completely described in the following with reference to the accompanying drawings in the embodiments of the present invention. It is obvious that the described embodiments are only a part of the embodiments of the present invention, but not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts are within the scope of the present invention.
参照图1所示，本实施例中公开了一种基于视频分析模式的跟踪统计人流量的系统，硬件方面主要包括：摄像头，存储装置，CPU计算装置等。在功能上包括：存储模块，提取模块，对比模块，定位模块，轨迹形成模块，方向模块，计数模块，计算模块，场景划分模块。Referring to FIG. 1 , a system for tracking statistic traffic based on a video analysis mode is disclosed in the embodiment. The hardware mainly includes a camera, a storage device, a CPU computing device, and the like. The functions include: a storage module, an extraction module, a comparison module, a positioning module, a track formation module, a direction module, a counting module, a calculation module, and a scene division module.
摄像头用于获取图像信息，摄像头设置位置根据具体场景设置，在存储模块中存储 预先建立的人体识别模型，并且在存储模块还存储预先自定义设置的运动轨迹；对比模块将摄像头获取的图像信息与存储模块中存储的人体识别模型进行对比；在对比之前，先对人体轮廓进行提取，提取模块从摄像头获取的图像信息中，将人体轮廓提取出来，并且传输至对比模块，对比模块将提取模块提取的人体轮廓与存储的人体识别模型进行对比。The camera is used to obtain image information, and the camera setting position is set according to a specific scene and stored in the storage module. a pre-established human body recognition model, and storing a motion trajectory of the pre-customized setting in the storage module; the comparison module compares the image information acquired by the camera with the human body recognition model stored in the storage module; before the comparison, the human body contour is first performed The extraction and extraction module extracts the human body contour from the image information acquired by the camera, and transmits the contour to the comparison module, and the comparison module compares the human body contour extracted by the extraction module with the stored human body recognition model.
对比模块在将提取的人体轮廓与存储的人体识别模块进行对比时，还包括对人体轮廓对比相似度阈值分析，当人体轮廓与存储的人体识别模型对比相似度高于阈值时，人体轮廓与人体识别模型匹配成功，当人体轮廓与存储的人体识别模块对比相似度低于阈值时，人体轮廓与人体识别模型匹配失败，不能将摄像头获取的图像定义为人。The contrast module compares the extracted human body contour with the stored human body recognition module, and also includes a threshold analysis of the human body contour contrast similarity. When the human body contour and the stored human body recognition model have a similarity similar to the threshold value, the human body contour and the human body contour When the recognition model is successfully matched, when the similarity between the human body contour and the stored human body recognition module is lower than the threshold value, the human body contour fails to match the human body recognition model, and the image acquired by the camera cannot be defined as a person.
对比模块进行人体识别模块对比后，如果获取的图像信息与人体识别模型对比成功，进行第一步定位；轨迹形成模块基于定位模块的第一步定位，对识别后的人体图像信息进行跟踪，形成三维空间中的人体移动轨迹。After comparing the human body recognition module with the comparison module, if the acquired image information is successfully compared with the human body recognition model, the first step is performed; the trajectory forming module tracks the human body image information after the recognition based on the first step of the positioning module to form The movement of the human body in three-dimensional space.
具体的，轨迹形成模块将进入方向定义为X，将出去方向定义为Y，通过定位模块进行第一步定位后，分别通过在X、Y两个方向上轨迹的减少或增加，来计算得到移动轨迹。Specifically, the trajectory forming module defines the entering direction as X and the outgoing direction as Y. After the first positioning by the positioning module, the movement is calculated by reducing or increasing the trajectory in the X and Y directions respectively. Track.
轨迹形成模块还对人体移动轨迹进行平滑度分析，判断平滑度是否满足预先自定义轨迹的阈值，如果满足，保留人体移动轨迹，如果不满足，则放弃人体移动轨迹。The trajectory forming module also performs smoothness analysis on the trajectory of the human body to determine whether the smoothness satisfies the threshold of the pre-custom trajectory. If it is satisfied, the trajectory of the human body is retained, and if not, the trajectory of the human body is discarded.
方向模块将轨迹形成模块形成的人体移动轨迹与存储模块中预先设置的运动轨迹做对比，确定人体移动轨迹的方向；然后计数模块根据方向模块确定的人体移动轨迹方向统计人流量；计算模块对计数模块统计的进入人流量和输出人流量进行统计计算，然后与后台系统数据进行分析，根据销售额和销售单数进行计算，得到人流转化率。计算模块针对人流量统计情况形成报表，根据报表后续分析店铺经营状况。The direction module compares the movement track of the human body formed by the track forming module with the preset motion track in the storage module to determine the direction of the human body moving track; then the counting module counts the human flow according to the direction of the human body moving track determined by the direction module; The module counts the incoming traffic and the output traffic for statistical calculation, and then analyzes with the background system data, and calculates according to the sales amount and the sales order number to obtain the conversion rate of the person flow. The calculation module forms a report for the statistics of the flow of people, and analyzes the operation status of the store according to the report.
在本实施例中，场景划分模块对图像中的检测区域进行场景划分，通过场景划分，获取检测区域内人体图像尺寸的变化范围。In this embodiment, the scene dividing module performs scene division on the detection area in the image, and obtains a variation range of the size of the human body image in the detection area by scene division.
本实施例中的原理：The principle in this embodiment:
先预先建立的人体识别模型和自定义设置的运动轨迹，摄像头在场景划分模块中获取人体图像，然后对人体轮廓进行提取，提取模块从摄像头获取的图像信息中，将人体轮廓提取出来，并且传输至对比模块，对比模块将提取模块提取的人体轮廓与存储的人体识别模型进行对比当人体轮廓与存储的人体识别模型对比相似度高于阈值时，人体轮廓与人体识别模型匹配成功，当人体轮廓与存储的人体识别模块对比相似度低于阈值时，人体轮廓与人体识别模型匹配失败，不能将摄像头获取的图像定义为人；对比模块进行人体识别模块对比后，如果获取的图像信息与人体识别模型对比成功，进行第一步定位；轨迹形成 模块基于定位模块的第一步定位，对识别后的人体图像信息进行跟踪，形成三维空间中的人体移动轨迹；对人体移动轨迹进行平滑度分析，判断平滑度是否满足预先自定义轨迹的阈值，如果满足，保留人体移动轨迹，如果不满足，则放弃人体移动轨迹；方向模块将轨迹形成模块形成的人体移动轨迹与存储模块中预先设置的运动轨迹做对比，确定人体移动轨迹的方向；然后计数模块根据方向模块确定的人体移动轨迹方向统计人流量。The pre-established human body recognition model and the custom set motion trajectory, the camera acquires the human body image in the scene division module, and then extracts the human body contour, and the extraction module extracts the human body contour from the image information acquired by the camera, and transmits To the comparison module, the comparison module compares the human body contour extracted by the extraction module with the stored human body recognition model. When the similarity between the human body contour and the stored human body recognition model is higher than the threshold value, the human body contour and the human body recognition model are successfully matched when the human body contour is When the similarity between the human body recognition module and the stored human body recognition module is lower than the threshold value, the human body contour fails to match the human body recognition model, and the image acquired by the camera cannot be defined as a person; the comparison module performs the human body recognition module comparison, if the acquired image information and the human body recognition model Contrast success, first step positioning; track formation The module is based on the first step of the positioning module to track the recognized human body image information to form a human body moving trajectory in a three-dimensional space; perform smoothness analysis on the human body moving trajectory to determine whether the smoothness satisfies the threshold of the pre-customized trajectory, If satisfied, the human body movement track is retained, and if not satisfied, the human body moving track is discarded; the direction module compares the human body moving track formed by the track forming module with the preset motion track in the storage module to determine the direction of the human body moving track; The module counts the flow of people according to the direction of the human body movement path determined by the direction module.
对所公开的实施例的上述说明，使本领域专业技术人员能够实现或使用本发明。对这些实施例的多种修改对本领域的专业技术人员来说将是显而易见的，本文中所定义的一般原理可以在不脱离本发明的精神或范围的情况下，在其它实施例中实现。因此，本发明将不会被限制于本文所示的这些实施例，而是要符合与本文所公开的原理和新颖特点相一致的最宽的范围。 The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments are obvious to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but the scope of the invention is to be accorded

Claims (7)

  1. 一种基于视频分析技术的跟踪统计人流量的系统，其特征在于，其包括：A system for tracking statistic traffic based on video analytics technology, characterized in that it comprises:
    摄像头，所述摄像头获取图像信息；a camera that acquires image information;
    存储模块，所述存储模块存储预先建立的人体识别模型，所述存储模块还存储预先自定义设置的运动轨迹；a storage module, the storage module stores a pre-established human body recognition model, and the storage module further stores a motion track preset in advance;
    对比模块，所述对比模块将所述摄像头获取的图像信息与所述存储模块中存储的人体识别模型进行对比；a comparison module, the comparison module compares image information acquired by the camera with a human body recognition model stored in the storage module;
    定位模块，所述对比模块进行人体识别模块对比后，如果获取的图像信息与人体识别模型对比成功，进行第一步定位；a positioning module, after the comparison module performs the comparison of the human body recognition module, if the acquired image information is successfully compared with the human body recognition model, the first step is performed;
    轨迹形成模块，所述轨迹形成模块基于所述定位模块的第一步定位，对识别后的人体图像信息进行跟踪，形成三维空间中的人体移动轨迹；a trajectory forming module, wherein the trajectory forming module tracks the recognized human body image information based on the first step of the positioning module to form a human body moving trajectory in a three-dimensional space;
    方向模块，所述方向模块将所述轨迹形成模块形成的人体移动轨迹与所述存储模块中预先设置的运动轨迹做对比，确定人体移动轨迹的方向；a direction module, the direction module compares a human body moving track formed by the track forming module with a preset motion track in the storage module, and determines a direction of a human body moving track;
    计数模块，所述计数模块根据所述方向模块确定的人体移动轨迹方向统计人流量；a counting module, wherein the counting module counts the flow of the human body according to the direction of the human body moving track determined by the direction module;
    计算模块，所述计算模块对所述计数模块统计的进入人流量和输出人流量进行统计计算，然后与后台系统数据进行分析，根据销售额和销售单数进行计算，得到人流转化率。a calculation module, the calculation module performs statistical calculation on the incoming traffic and the output traffic calculated by the counting module, and then analyzes with the background system data, and performs calculation according to the sales amount and the sales order number to obtain the conversion rate of the human flow.
  2. 根据权利要求1所述的一种基于视频分析技术的跟踪统计人流量的系统，其特征在于，还包括提取模块，所述提取模块从所述摄像头获取的图像信息中，将人体轮廓提取出来，并且传输至所述对比模块，所述对比模块将所述提取模块提取的人体轮廓与存储的人体识别模型进行对比。The system for tracking statistic person traffic based on video analysis technology according to claim 1, further comprising an extraction module, wherein the extraction module extracts a contour of a human body from image information acquired by the camera, And transmitting to the comparison module, the comparison module compares the human body contour extracted by the extraction module with the stored human body recognition model.
  3. 根据权利要求2所述的一种基于视频分析技术的跟踪统计人流量的系统，其特征在于，所述对比模块在将提取的人体轮廓与存储的人体识别模块进行对比时，还包括对人体轮廓对比相似度阈值分析，当人体轮廓与存储的人体识别模型对比相似度高于阈值时，人体轮廓与人体识别模型匹配成功，当人体轮廓与存储的人体识别模块对比相似度低于阈值时，人体轮廓与人体识别模型匹配失败，不能将所述摄像头获取的图像定义为人。The system for tracking statistic human traffic based on video analysis technology according to claim 2, wherein the comparison module further comprises a contour of the human body when comparing the extracted human body contour with the stored human body recognition module Comparing the similarity threshold analysis, when the similarity between the human body contour and the stored human body recognition model is higher than the threshold value, the human body contour and the human body recognition model are successfully matched. When the human body contour and the stored human body recognition module are less than the threshold value, the human body contour is compared. The contour fails to match the human body recognition model, and the image acquired by the camera cannot be defined as a person.
  4. 根据权利要求1所述的一种基于视频分析技术的跟踪统计人流量的系统，其特征在于，所述轨迹形成模块将进入方向定义为X，将出去方向定义为Y，通过所述定位模块进行第一步定位后，分别通过在X、Y两个方向上轨迹的减少或增加，来计算得到移动轨迹。The system for tracking statistician traffic based on video analysis technology according to claim 1, wherein the trajectory forming module defines an entry direction as X and an outgoing direction as Y, by using the positioning module. After the first step of positioning, the movement trajectory is calculated by reducing or increasing the trajectory in both the X and Y directions.
  5. 根据权利要求4所述的一种基于视频分析技术的跟踪统计人流量的系统，其特征在 于，所述轨迹形成模块还对人体移动轨迹进行平滑度分析，判断平滑度是否满足预先自定义轨迹的阈值，如果满足，保留人体移动轨迹，如果不满足，则放弃人体移动轨迹。A system for tracking statistic traffic based on video analysis technology according to claim 4, characterized in that The trajectory forming module further performs smoothness analysis on the trajectory of the human body to determine whether the smoothness satisfies the threshold of the pre-custom trajectory. If it is satisfied, the trajectory of the human body is retained, and if not, the trajectory of the human body is discarded.
  6. 根据权利要求1所述的一种基于视频分析技术的跟踪统计人流量的系统，其特征在于，还包括场景划分模块，所述场景划分模块对图像中的检测区域进行场景划分，通过场景划分，获取检测区域内人体图像尺寸的变化范围。The system for tracking statistician traffic based on the video analysis technology of claim 1, further comprising a scene division module, wherein the scene division module performs scene division on the detection area in the image, and divides the scene by Obtain the range of variation of the size of the human body image in the detection area.
  7. 根据权利要求1所述的一种基于视频分析技术的跟踪统计人流量的系统，其特征在于，所述计算模块针对人流量统计情况形成报表，根据报表后续分析店铺经营状况。 The system for tracking statistic traffic based on video analysis technology according to claim 1, wherein the calculation module forms a report for the statistics of the human traffic, and analyzes the operation status of the store according to the report.
PCT/CN2017/111929 2016-12-30 2017-11-20 System for collecting statistics on pedestrian traffic by means of tracking based on video analysis technique WO2018121127A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201611256466.3A CN106650695A (en) 2016-12-30 2016-12-30 Video analysis technology-based people flow tracking statistics system
CN201611256466.3 2016-12-30

Publications (1)

Publication Number Publication Date
WO2018121127A1 true WO2018121127A1 (en) 2018-07-05

Family

ID=58838572

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/111929 WO2018121127A1 (en) 2016-12-30 2017-11-20 System for collecting statistics on pedestrian traffic by means of tracking based on video analysis technique

Country Status (2)

Country Link
CN (1) CN106650695A (en)
WO (1) WO2018121127A1 (en)

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109583411A (en) * 2018-12-09 2019-04-05 大连海事大学 The online checking method of tourist's classification based on TOF camera
CN110175562A (en) * 2019-05-24 2019-08-27 深圳职业技术学院 A kind of gpu acceleration pedestrian detection algorithm based on parallel computation
CN111860261A (en) * 2020-07-10 2020-10-30 北京猎户星空科技有限公司 Passenger flow value statistical method, device, equipment and medium
CN111914754A (en) * 2020-08-03 2020-11-10 杭州云栖智慧视通科技有限公司 Image content similarity measurement method and device and computer equipment
CN113051975A (en) * 2019-12-27 2021-06-29 深圳云天励飞技术有限公司 People flow statistical method and related product
CN113326830A (en) * 2021-08-04 2021-08-31 北京文安智能技术股份有限公司 Passenger flow statistical model training method and passenger flow statistical method based on overlook images
CN115215177A (en) * 2022-08-26 2022-10-21 中建八局发展建设有限公司 Intelligent identification system and method for elevator lifting
CN116542858A (en) * 2023-07-03 2023-08-04 众芯汉创(江苏)科技有限公司 Data splicing analysis system based on space track
CN117226843A (en) * 2023-09-27 2023-12-15 盐城工学院 Robot movement track control method and system based on visual servo

Families Citing this family (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106650695A (en) * 2016-12-30 2017-05-10 苏州万店掌网络科技有限公司 Video analysis technology-based people flow tracking statistics system
CN107358163A (en) * 2017-06-14 2017-11-17 广东数相智能科技有限公司 Visitor's line trace statistical method, electronic equipment and storage medium based on recognition of face
CN108090454A (en) * 2017-12-26 2018-05-29 上海理工大学 Campus bathhouse people flow rate statistical system
CN108833420A (en) * 2018-06-23 2018-11-16 苏州万店掌网络科技有限公司 A kind of system based on video analysis and wifi probe combinations analysis passenger flow
CN108875677B (en) * 2018-06-29 2020-11-17 上海小蚁科技有限公司 Passenger flow volume statistical method and device, storage medium and terminal
CN109902551A (en) * 2018-11-09 2019-06-18 阿里巴巴集团控股有限公司 The real-time stream of people's statistical method and device of open scene
CN110659578A (en) * 2019-08-26 2020-01-07 中国电子科技集团公司电子科学研究院 Passenger flow statistical method, system and equipment based on detection and tracking technology
CN112580397A (en) * 2019-09-29 2021-03-30 北京市商汤科技开发有限公司 Data processing method, device and storage medium
JP2022503373A (en) * 2019-09-29 2022-01-12 北京市商▲湯▼科技▲開▼▲發▼有限公司 Data processing methods, devices and storage media
CN112001941B (en) * 2020-06-05 2023-11-03 成都睿畜电子科技有限公司 Piglet supervision method and system based on computer vision
CN112668525B (en) * 2020-12-31 2024-05-07 深圳云天励飞技术股份有限公司 People flow counting method and device, electronic equipment and storage medium
CN113369796A (en) * 2021-05-26 2021-09-10 上海御建装饰设计工程有限公司 Intelligent automatic tracking and monitoring system for cultural relic restoration
CN113743324B (en) * 2021-09-07 2022-10-18 易科捷(武汉)生态科技有限公司成都分公司 Automatic updating type fish identification system based on Internet of things

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2001018697A1 (en) * 1999-09-06 2001-03-15 Holger Lausch Method and device for detecting and analyzing the reception behavior of people
CN101587539A (en) * 2008-05-21 2009-11-25 上海新联纬讯科技发展有限公司 Method and system of population flow statistics based on intelligent video identification technology
CN101937534A (en) * 2010-09-17 2011-01-05 上海沐盈信息科技有限公司 Electronic marketing management system and management method
US7987111B1 (en) * 2006-10-30 2011-07-26 Videomining Corporation Method and system for characterizing physical retail spaces by determining the demographic composition of people in the physical retail spaces utilizing video image analysis
CN202374374U (en) * 2011-10-25 2012-08-08 武汉金数字图像信号设备有限公司 Shopping mall crowd flow dynamic management system based on remote images
CN102750539A (en) * 2012-06-06 2012-10-24 深圳市海亿达能源科技股份有限公司 Device and method for counting regional space population
CN103425967A (en) * 2013-07-21 2013-12-04 浙江大学 Pedestrian flow monitoring method based on pedestrian detection and tracking
CN106650695A (en) * 2016-12-30 2017-05-10 苏州万店掌网络科技有限公司 Video analysis technology-based people flow tracking statistics system

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101739653A (en) * 2008-11-13 2010-06-16 上海汇纳网络信息科技有限公司 Passenger flow analysis system
CN102982598B (en) * 2012-11-14 2015-05-20 三峡大学 Video people counting method and system based on single camera scene configuration
CN103093274B (en) * 2012-12-30 2016-06-29 信帧电子技术（北京）有限公司 Method based on the people counting of video
US20160019698A1 (en) * 2014-07-21 2016-01-21 Florida Atlantic University Systems and methods for people counting in sequential images

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2001018697A1 (en) * 1999-09-06 2001-03-15 Holger Lausch Method and device for detecting and analyzing the reception behavior of people
US7987111B1 (en) * 2006-10-30 2011-07-26 Videomining Corporation Method and system for characterizing physical retail spaces by determining the demographic composition of people in the physical retail spaces utilizing video image analysis
CN101587539A (en) * 2008-05-21 2009-11-25 上海新联纬讯科技发展有限公司 Method and system of population flow statistics based on intelligent video identification technology
CN101937534A (en) * 2010-09-17 2011-01-05 上海沐盈信息科技有限公司 Electronic marketing management system and management method
CN202374374U (en) * 2011-10-25 2012-08-08 武汉金数字图像信号设备有限公司 Shopping mall crowd flow dynamic management system based on remote images
CN102750539A (en) * 2012-06-06 2012-10-24 深圳市海亿达能源科技股份有限公司 Device and method for counting regional space population
CN103425967A (en) * 2013-07-21 2013-12-04 浙江大学 Pedestrian flow monitoring method based on pedestrian detection and tracking
CN106650695A (en) * 2016-12-30 2017-05-10 苏州万店掌网络科技有限公司 Video analysis technology-based people flow tracking statistics system

Cited By (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109583411A (en) * 2018-12-09 2019-04-05 大连海事大学 The online checking method of tourist's classification based on TOF camera
CN109583411B (en) * 2018-12-09 2022-10-21 大连海事大学 TOF camera-based tourist category online auditing method
CN110175562A (en) * 2019-05-24 2019-08-27 深圳职业技术学院 A kind of gpu acceleration pedestrian detection algorithm based on parallel computation
CN110175562B (en) * 2019-05-24 2024-01-16 深圳职业技术大学 Parallel computing-based gpu acceleration pedestrian detection algorithm
CN113051975A (en) * 2019-12-27 2021-06-29 深圳云天励飞技术有限公司 People flow statistical method and related product
CN113051975B (en) * 2019-12-27 2024-04-02 深圳云天励飞技术有限公司 People flow statistics method and related products
CN111860261B (en) * 2020-07-10 2023-11-03 北京猎户星空科技有限公司 Passenger flow value statistical method, device, equipment and medium
CN111860261A (en) * 2020-07-10 2020-10-30 北京猎户星空科技有限公司 Passenger flow value statistical method, device, equipment and medium
CN111914754A (en) * 2020-08-03 2020-11-10 杭州云栖智慧视通科技有限公司 Image content similarity measurement method and device and computer equipment
CN113326830A (en) * 2021-08-04 2021-08-31 北京文安智能技术股份有限公司 Passenger flow statistical model training method and passenger flow statistical method based on overlook images
CN115215177A (en) * 2022-08-26 2022-10-21 中建八局发展建设有限公司 Intelligent identification system and method for elevator lifting
CN116542858B (en) * 2023-07-03 2023-09-05 众芯汉创(江苏)科技有限公司 Data splicing analysis system based on space track
CN116542858A (en) * 2023-07-03 2023-08-04 众芯汉创(江苏)科技有限公司 Data splicing analysis system based on space track
CN117226843A (en) * 2023-09-27 2023-12-15 盐城工学院 Robot movement track control method and system based on visual servo
CN117226843B (en) * 2023-09-27 2024-02-27 盐城工学院 Robot movement track control method and system based on visual servo

Also Published As

Publication number Publication date
CN106650695A (en) 2017-05-10

Similar Documents

Publication Publication Date Title
WO2018121127A1 (en) System for collecting statistics on pedestrian traffic by means of tracking based on video analysis technique
Bondi et al. Real-time people counting from depth imagery of crowded environments
Luber et al. People tracking in rgb-d data with on-line boosted target models
US8582816B2 (en) Method and apparatus for video analytics based object counting
Zaki et al. Automated analysis of pedestrian group behavior in urban settings
WO2015131734A1 (en) Method, device, and storage medium for pedestrian counting in forward looking surveillance scenario
Porikli Trajectory distance metric using hidden markov model based representation
Ryan et al. Crowd counting using group tracking and local features
US20130314505A1 (en) System And Process For Detecting, Tracking And Counting Human Objects of Interest
US20130243252A1 (en) Loitering detection in a video surveillance system
Jodoin et al. Tracking all road users at multimodal urban traffic intersections
Kang et al. Tracking people in crowded scenes across multiple cameras
Liu et al. Real-time human detection and tracking in complex environments using single RGBD camera
WO2008092393A1 (en) Method of moving target tracking and number accounting
Stahlschmidt et al. Applications for a people detection and tracking algorithm using a time-of-flight camera
Mokayed et al. Real-time human detection and counting system using deep learning computer vision techniques
US9965687B2 (en) System and method for detecting potential mugging event via trajectory-based analysis
Miller et al. A hidden Markov model for vehicle detection and counting
Pore et al. Bidirectional people counting system in video surveillance
Kocamaz et al. Vision-based counting of pedestrians and cyclists
Virgona et al. Sensing and perception technology to enable real time monitoring of passenger movement behaviours through congested rail stations
Kokul et al. Online multi-person tracking-by-detection method using ACF and particle filter
Felemban et al. Deep Trajectory Classification Model for Congestion Detection in Human Crowds.
Heili et al. Parameter estimation and contextual adaptation for a multi-object tracking CRF model
Zhang et al. A robust human detection and tracking system using a human-model-based camera calibration

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17888231

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17888231

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS (EPO FORM 1205A DATED 04.12.19)

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS (EPO FORM 1205A DATED 04.12.2019)

122 Ep: pct application non-entry in european phase

Ref document number: 17888231

Country of ref document: EP

Kind code of ref document: A1