Dense crowd abnormal behavior detection method based on micro-behavior analysis
A technology of dense crowds and detection methods, applied in instruments, character and pattern recognition, computer parts, etc., to achieve the effect of abnormal behavior detection, avoiding tracking and detection problems, and high accuracy
Inactive Publication Date: 2014-02-19
CHINA JILIANG UNIV
View PDF3 Cites 24 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
For this field, the method proposed by S. Ali et al is to use crowd flow to help track a single target, see Floor fields for tracking in high density crowd scenes, in ECCV '08: Proceedings of the 10th European Conference on Computer Vision, 2008, pp. 1–14., but this method is only for dense crowds moving at a uniform speed
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0026] refer to figure 1 , the specific implementation process of the present invention is as follows:
[0027] One: Input the video sequence, according to the size of the person in the video, divide the original video image into blocks, Represents the number of horizontal partition blocks, Represents the number of vertically divided blocks, the central pixel of each image block is used as a sampling point, and the composition size is sampled image. The segmentation of image blocks is determined by the space occupied by people in the video. The purpose is to make the information of each sampling point reflect the different individual behavior information in the actual scene of dense crowd as much as possible. It can be seen that the higher the crowd density, the more sampling points , the calculation is more accurate. Therefore, our method is suitable for both large and small groups of people. Obtained by the above method sampling points.
[0028] Two: Dynamical s...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses a dense crowd abnormal behavior detection method based on micro-behavior analysis. The dense crowd abnormal behavior detection method comprises the steps that according to sizes of people in a dense video, image sampling is carried out on an original video according to a proportion, and sampling images are obtained; a dynamic system, defined by particle flows, of the sampling images is calculated; the speed and position information of the dynamic system are utilized to carry out micro-behavior analysis on sampling particles; a sample sequence is constructed, information extraction is carried out on micro-behaviors of the particles, the information is quantized and marked, and a space-time cube is constructed; dense crowd video sequences for training are collected, sample feature sequences of the video sequences are extracted through the steps and serve a observation sequences for hidden conditional random field training, and hidden conditional random fields are utilized to achieve dense crowd behavior detection. The dense crowd abnormal behavior detection method aims at resolving the problem that crowd behavior understanding has potential risks due to the fact that single-object detection and tracking cannot be efficiently achieved in a high-density environment, is suitable for scenes with complex crowd behaviors, and is used for identifying specific crowd behaviors and carrying out anomaly detection.
Description
technical field [0001] The invention belongs to the field of behavior recognition of dense crowds in videos, and in particular relates to a method for identifying common behaviors of dense crowds based on micro-behavior analysis and abnormal processing. The method can be applied to crowd flow comparisons at stations, shopping malls, and emergency entrances and exits big place. Background technique [0002] Video analysis of dense crowd scenes poses a huge challenge to computer vision techniques. The high-density environment in real life makes the identification and tracking of individual objects impractical. Understanding the behavior of dense crowds without knowing the behavior of individuals is very dangerous. The intelligent detection technology of dense crowd behavior has a wide range of applications. For example, it can predict the congestion situation and avoid the occurrence of tragedies. [0003] At present, video surveillance technology (applied to stations, road...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More IPC IPC(8): G06K9/00G06K9/62
Inventor 章东平徐娇李世忠彭怀亮
Owner CHINA JILIANG UNIV