Image processing method and device based on anchor point image cutting, equipment and medium

Abstract

The embodiment of the invention discloses an image processing method and device based on anchor point image cutting, equipment and a medium, and the method comprises the steps: obtaining the anchor point information of an image based on the annotation information of the image; wherein the anchor point information may include center point information and size information of the anchor point identification frame. The size information of the sliding window can be accurately determined according to the set maximum value of the sliding window, the minimum value of the sliding window, the size multiple and the size information of the anchor point identification frame; and on the basis of the sliding window size information and the center point information, a cut image containing the target object can be intercepted from the image, the initial depth detection network model is trained by using the cut image, and a depth detection network model used for detecting the position of the target object in the image is obtained. For a super-large image, the image can be intercepted into a cut image with a relatively small pixel value in an image cutting manner, so that the initial depth detection network model can realize detection and analysis of the cut image, and the purpose of detection and analysis of the super-large image is achieved.

Description

technical field [0001] The present application relates to the technical field of computer vision, and in particular relates to an image processing method, device, device and computer-readable storage medium based on anchor point cutting. Background technique [0002] Object detection is one of the most important research directions in the field of computer vision. Due to its strong practicability and landing prospects, a large number of researchers have joined the research community for target algorithm optimization. The target detection algorithm is gradually becoming mature, and various algorithms and models are emerging. At this stage, due to the popularity of deep learning and the general improvement of computing power, cutting-edge target detection algorithms such as Yolo v5 and the target detection network (Center Net) model have achieved good performance in terms of efficiency and accuracy. Among them, Yolo v5 is a fast and compact open source object detection model...

AI Technical Summary

Problems solved by technology

[0004] Therefore, existing detection algorithms cannot solve the detection problem of super-large pixel images

On the one hand, the computing power of existing algorithms cannot meet the requirements of inputting the entire image into the algorithm model for reasoning; on the other hand, simple scaling of the image will lose a large number of detailed features, resulting in serious missed detection of small objects

Images