Human body identifying and tracking method

Abstract

The invention discloses a human body recognition tracking method, which comprises the following steps: the step 100, acquiring original video stream data, and converting the original video stream data into pictures to establish an initial data set; the step 200, performing enhancement processing and screening on the initial data set to obtain a training set, a verification set and a test set; the step 300, building a Centernet network structure composed of a backbone network, an up-mining path and a top convolution, wherein the top convolution adopts a deep separable convolution; the step 400, designing a BOX matching mechanism and a loss function to construct a complete Centernet network structure; the step 500, using the training set, the verification set and the test set to train, verify and test the complete Centernet network structure to obtain a Centernet network model; and the step 600, identifying and tracking the human body in the real-time video stream data by using the Centernet network model. According to the human body identification tracking method, the Centernet network structure is optimized, the detection speed is improved under the condition that the detection accuracy is not reduced, and the balance between the accuracy and the speed is optimized.

Description

technical field [0001] The invention relates to the field of machine vision, in particular to a human body recognition and tracking method. Background technique [0002] Multi-Object Tracking (Multi-Object Tracking, MOT) is a current research hotspot in the field of computer vision. Information such as the complete trajectory of a target. In recent years, with the rapid growth of data processing capabilities and the development of image analysis technology, target monitoring and real-time tracking technologies have come to the fore, and have very important practical value in the fields of video surveillance, positioning and navigation, intelligent human-computer interaction, virtual reality, etc. Multi-target tracking technology based on video stream has become a hot research direction of various experts and scholars. [0003] As a target tracking algorithm, the Centernet network does not require a region to establish a region of interest, and the speed has been greatly im...

AI Technical Summary

Problems solved by technology

[0003] As a target tracking algorithm, the Centernet network does not require an area to establish an area of ​​interest, and the speed has been greatly improved, but there is still room for optimization in the balance between detection accuracy and detection speed

Images