A Generation Method of Optimal Object Detection Scale in People Flow Statistics

Abstract

The invention discloses a method for generating an optimal target detection scale in people flow statistics, comprising the following steps: acquiring a video source, and initializing the statistical target queue people_list to be empty; reading the current frame image of the video source, and converting the image into gray degree image Gk, and initialize the detection target queue head_list to be empty, divide the grayscale image Gk into n regions; perform target detection on the grayscale image Gk, update the people_list queue by using the detection target and tracking information, and new unrelated The detection target is added to the people_list as a new target; after processing a certain number of video frames, calculate the target detection scale distribution histogram in each area according to the current position information of all statistical targets in the people_list, and use the target detection scale distribution histogram information to calculate the optimal detection in each area scale. The invention can have beneficial effects on the high detection rate and low false alarm rate of the video-based people flow algorithm.

Description

technical field [0001] The invention belongs to the field of computer vision, and more specifically relates to a method for generating an optimal target detection scale in people flow statistics. Background technique [0002] People flow counting technology is widely used in shopping malls, subways, stations and other places, and the research on this technology also has extremely high market value. The traditional way of counting the flow of people mostly adopts manual counting method, which needs to hire a lot of manpower, resulting in a waste of resources. With the development of computer vision technology, people flow counting technology based on video images is gradually developing. It automatically processes and analyzes the video images acquired by the camera by using relevant theories in image processing and other disciplines to analyze the images in the images. The target is identified and tracked, so as to realize the function of automatic counting. This method sa...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the above defects or improvement needs of the prior art, the present invention provides a method for generating the optimal target detection scale in people flow statistics, the purpose of which is to avoid the inconvenience of manually setting the target detection scale in the video-based people flow statistics algorithm. Accuracy and inconvenience issues, and can have a beneficial effect on the high detection rate and low false alarm rate of video-based people traffic algorithms

Images