Portable method, system and device for marking target to be identified, and storage medium

Abstract

The invention provides a portable method, system and device for marking a target to be identified, and a storage medium. The portable method for marking a target to be identified includes the following steps: Step S1, collecting a plurality of images, each image at least including a commodity area; Step S2, presetting a marking box, so that the image can move relative to the marking frame, so as to dragging the commodity area on each image to the marking box; and Step S3, marking the commodity area through the marking frame, so that each commodity area has a commodity marking box. According tothe method, the image is dragged, so that the commodity area on the image is dragged to a preset marking box to be subjected to marking of the commodity area, the possibility of collecting an image and marking the image through a mobile terminal such as a mobile phone is realized, the efficiency of image collection and marking is improved, complicated work of image marking can be outsourced through a mobile phone APP, thereby realizing the possibility of collection and marking of massive images.

Description

technical field [0001] The present invention relates to artificial intelligence, in particular to a convenient marking method, system, equipment and storage medium for an object to be identified. Background technique [0002] In the fast-moving consumer goods industry, in order to obtain the SKU (Stock Keeping Unit) of product information in a timely manner, it is necessary to collect the products on the shelf, and further target the products in the image. [0003] The traditional target marking method is an image-based computer vision target recognition method. First, the obtained product image is preprocessed to reduce noise, and then target recognition is performed, and template matching is applied or region of interest is extracted based on grayscale changes. However, the collected images often contain multiple categories of goods, and the packaging is similar, there are many types, and the placement is different. For a computer, it is relatively computationally intensiv...

AI Technical Summary

Problems solved by technology

[0004] The method based on artificial target identification is that the user uses the computer to mark, cut and other processes on the image. The commodities in the image are generally arranged in a list. Generally, the regions of interest are selected one by one and the selection operation is performed. When the user needs to process When there are too many pictures, the user needs to click one by one, the operation is extremely cumbersome, and at the same time, because the frequent clicking process may make the user lose concentration and cause misoperation, it takes a long time and the user experience is not good

Images