Image processing method and device, electronic equipment and storage medium

Abstract

The invention discloses an image processing method and device, electronic equipment and a storage medium. The method comprises the steps: obtaining a to-be-processed image through a single camera; preprocessing the to-be-processed image to obtain a preprocessed image; performing down-sampling operation on the preprocessed image based on a pre-trained convolutional neural network model to obtain afeature image; and performing up-sampling operation on the feature image to obtain a depth prediction image. The invention further provides an image processing device, electronic equipment and a storage medium. According to the invention, environmental depth-of-field prediction can be carried out through the single camera, and the accuracy and rate of environmental depth-of-field prediction are improved at the same time.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to an image processing method, device, electronic equipment and storage medium. Background technique [0002] With the development and progress of artificial intelligence technology, environmental depth prediction technology has been widely used in many cutting-edge technologies, such as: automatic driving, 3D reconstruction of 2D scenes, etc. Environmental depth prediction technology is one of the foundations of many scientific research and commercial products one. [0003] The traditional environment depth prediction method is to imitate the left and right eyes of a person to take two photos of the same scene from different angles through a binocular camera. Because there are differences in the same object in photos taken from different angles, this difference is called "parallax". Usually, the smaller the parallax, the closer the object is to the camera, and th...

AI Technical Summary

Problems solved by technology

This process usually requires camera calibration, image correction, stereo matching and other processes, and the calculation amount is relatively large. At the same time, traditional environmental depth prediction is mostly based on dual cameras, multi-cameras or laser ranging, which brings high system costs. , High requirements for system computing power, inability to miniaturize, etc.

Images