Commodity touch information clustering method, device and equipment based on surveillance video

Abstract

The invention relates to a commodity touch information clustering method, device and equipment based on a surveillance video. The method comprises the following steps of: extracting video images frameby frame from a real-time surveillance video of a commodity; generating a differential gradient image and a binary feature image of a detection area based on a current video image and a preset background image; taking the differential gradient image as a detection image if the number of pixel points of which the pixel values are not zero in the binary feature image is greater than a preset threshold value; counting the touched times of different commodities through an information clustering method based on the detection image; and displaying the touched times in the real-time surveillance video. According to the method, the number of touched times of the commodities in the specific area is recognized in an information clustering mode, so that the attention degree of the commodities in thearea in the market can be reflected, and the method is simple in detection algorithm, small in calculated amount, robust in technology and good in detection effect.

Description

technical field [0001] The present application relates to the technical field of computer vision, and in particular to a surveillance video-based product touch information clustering method, device and equipment. Background technique [0002] At present, in retail stores of various commodities, cameras are usually set up to monitor the commodities sold. One is to avoid theft and losses. Second, merchants can also use surveillance videos to know which commodities are more concerned, that is, which commodities have been touched. The number of touches is more, thereby improving the accuracy of daily store management, merchandising strategies, and stocking plans. [0003] For the above-mentioned second use, there are currently several different implementation methods. One is the artificial statistical method, that is, the artificial naked eye continues to watch the surveillance video, and its disadvantage is that the labor cost is high; the other is the image-based object detect...

AI Technical Summary

Problems solved by technology

[0004] This application provides a surveillance video-based product touch information clustering method, device, and equipment to solve the problems of high labor costs or complicated detection processes in existing product touch detection methods

Images