Picture segmentation method and device, storage medium and computer equipment

Abstract

The invention provides a picture segmentation method. The method comprises that a picture to be segmented is obtained, a convolutional kernel is used to carry out edge extraction on the picture to besegmented to obtain an edge feature map, a corresponding edge binary image is obtained according to the edge feature map, straight line analysis is carried out on the edge binary image to obtain a straight line analysis formula corresponding to the edge binary image, and candidate straight line combinations that can form a polygon are determined according to the straight line analysis formula; candidate parameter information is calculated according to geometric information of the polygon and geometric information of the picture to be segmented, and the candidate parameter information matches standard parameter information to screen a target polygon; and a target straight line combination corresponding to the target polygon is obtained from the candidate straight line combinations, and a target segmentation area of the picture to be segmented is determined according to the target straight line analysis formula corresponding to the target straight line combination. Operation is more convenient via accurate edge extraction and geometric analysis. The invention also provides a picture segmentation device, a storage medium and computer equipment.

Description

technical field [0001] The present invention relates to the field of image processing, in particular to a picture segmentation method, device, storage medium and computer equipment. Background technique [0002] With the continuous development of science and technology, the functions of mobile terminals are becoming more and more perfect. You can use mobile terminals to watch videos, instant chat, online shopping, etc. In the process of using various functions of mobile terminals, in order to obtain relevant information, you need to collect pictures Part of the information in the ID card, such as identity verification, first uses the camera to take a picture of the ID card, and then divides the picture containing the ID card to obtain the target area of ​​the ID card in the picture, and further identifies the identity information. [0003] When using a mobile terminal to identify picture information, it is necessary to first collect pictures or video streams through a camera...

AI Technical Summary

Problems solved by technology

Traditional technology has poor edge extraction ability, resulting in low recognition accuracy. When taking pictures, users need to keep the target object at a fixed position of the camera to ensure that the segmented area is located at a specific position of the picture, which is inconvenient to operate.

Images