Human body behavior recognition method based on history motion graph and R transformation

Abstract

The invention discloses a human body behavior recognition method based on a history motion graph and R transformation. According to the method, a depth video is used as a recognition basis, firstly, the minimum enclosure rectangle of human body motion is calculated according to a foreground segmentation technology, then the history motion graph is extracted within a depth video area limited by the minimum enclosure rectangle, motion intensity constraint is exerted on the extracted history motion graph, so that a motion energy diagram is obtained, R transformation is calculated on the obtained motion energy graph, and therefore a characteristic vector used for behavior recognition is obtained. A method of a support vector machine is adopted for training and recognition processes. The minimum enclosure rectangle of human body behavior motion is adopted for preprocessing, and behavior characteristic extraction is accelerated; a method of history motion graph sequences is adopted for reducing influences of noise in depth graphs; characteristics are extracted through performing R transformation on the energy graph, so that calculation speed is high.

Description

technical field [0001] The invention relates to the fields of computer vision and image processing, in particular to a human behavior recognition method based on a motion history graph and R transformation. Background technique [0002] Video surveillance is a hot and key issue in the field of visual research today. In the field of security and human-computer interaction, a large number of video data are continuously generated. These data are often measured in units of G. Manual judgment alone will undoubtedly It consumes a lot of manpower. Videos are rich in content. Most of the time we only pay attention to certain parts of the video, such as human behavior. If it can be automatically and efficiently identified, it will liberate a lot of manpower. The current behavior recognition research results mainly focus on the behavior recognition research of RGB video. [0003] RGB video is the most common form of video, with a wide range of sources. There have been many research r...

AI Technical Summary

Problems solved by technology

These methods have their own limitations: the A bag of 3D words method has a high recognition accuracy rate, but due to the need for uniform sampling on the human body contour, the depth data obtained is required to be very pure, and cannot be used in human behavior recognition in actual scenes; The method of directly applying 3D-MHIs is fast enough, but the recognition accuracy is not enough; DMM-HOG is also effective for behavior recognition in complex backgrounds while ensuring recognition accuracy, but this method is too time-consuming to achieve real-time human behavior identify

Images