Track generation method and device based on video image, electronic equipment and medium

Abstract

The embodiment of the invention discloses a track generation method and device based on a video image, electronic equipment and a medium. A specific embodiment of the method comprises the following steps: performing article entity detection on each frame of image in a pre-acquired video to generate article detection information and obtain an article detection information set; generating a candidate track set based on article detection information in the article detection information set; screening candidate tracks meeting track screening conditions out from the candidate track set to serve as pre-processing tracks, and obtaining a pre-processing track sequence; performing track fusion processing on the preprocessed tracks in the preprocessed track sequence to obtain a fused track set; based on the fusion track set, generating an article taking information group set; and sending the article taking information set to a target terminal for display. According to the embodiment, the space utilization rate of the unmanned container is increased.

Description

technical field [0001] The embodiments of the present disclosure relate to the field of computer technology, and in particular to a method, device, electronic device and medium for generating a trajectory based on video images. Background technique [0002] With the rapid development of smart technology, the way people shop has undergone tremendous changes. At the same time, the vending model based on unmanned containers has developed rapidly. As a kind of unmanned container, the dynamic container is usually identified by static items. That is, by identifying before and after the door of the unmanned vending cabinet is opened, the items in the cabinet change. In this way, the item actually taken by the user is determined, and then the picking information of the item is generated. However, when the above method is adopted, the following technical problems usually exist: [0003] First, using the static item identification method, in order to ensure that the changes of the...

AI Technical Summary

Problems solved by technology

[0003] First, using the static item identification method, in order to ensure that the changes of the items in the cabinet can be accurately determined, it is often necessary to reserve enough space in the unmanned container to ensure that the camera in the unmanned container can clearly capture the unmanned container All the items in the container, this method makes the space utilization rate of the unmanned container lower;

[0004] Second, due to problems such as light, the clarity of the video may change, and in turn, the detection model is used to detect the image in the video, and the position of the generated item is not accurate enough, so that it is impossible to accurately determine what the user picked up. Items, so that the generated item retrieval information is often wrong, which affects the user's shopping experience and may reduce the user flow of unmanned containers

Images