A filtering method, electronic equipment and storage medium of a prediction frame

Abstract

The embodiment of the present invention relates to the field of image processing, and discloses a filtering method of a prediction frame, an electronic device and a storage medium. In some embodiments of the present application, the filtering method of the prediction frame includes: determining the effective prediction frame and the prediction frame to be filtered in the distorted image; for each prediction frame to be filtered, perform the following operations respectively: according to the prediction frame to be filtered The degree of overlap with the effective prediction frame and the overlap threshold corresponding to the prediction frame to be filtered determine whether to keep the prediction frame to be filtered, wherein the overlap threshold is determined according to the reference position information corresponding to the prediction frame to be filtered. In this implementation, in the process of filtering the prediction frame, the overlap threshold is set for different prediction frames, which reduces the situation that the prediction frame in the distorted image is accidentally deleted or omitted compared with the method of using a fixed overlap threshold. .

Description

technical field [0001] The embodiments of the present invention relate to the field of image processing, and in particular to a method for filtering prediction frames, electronic equipment and storage media. Background technique [0002] A fisheye lens is a special lens with a wider angle of view and a shorter focal length than standard lenses. Due to the large viewing angle and short focal length of the fisheye lens, the captured fisheye images are distorted. At present, the industry mainly uses fisheye lenses as visual sensors to collect fisheye images. [0003] However, the inventors have found that at least the following problems exist in the prior art: at present, the accuracy rate of the deep learning target detection algorithm is already significantly higher than that of the traditional target detection and recognition algorithm. The accuracy of deep learning algorithms is greatly affected by the training data. When performing a deep learning target detection algor...

AI Technical Summary

Problems solved by technology

[0003] However, the inventors found that there are at least the following problems in the prior art: at present, the accuracy rate of the deep learning target detection algorithm is already significantly higher than that of the traditional target detection and recognition algorithm

Images