Image processing method, device and equipment and computer readable storage medium

Abstract

The embodiment of the invention discloses an image processing method, device and equipment and a computer readable storage medium. The method comprises the steps: obtaining a sample image and a labeldepth image corresponding to the sample image; inputting the sample image into a trained first image processing model to obtain a first depth estimation feature; inputting the sample image into a second image processing model to obtain a second depth estimation feature, and generating a predicted depth image corresponding to the sample image according to the second depth estimation feature, wherein the number of model parameters in the second image processing model is smaller than the number of model parameters in the first image processing model; and adjusting model parameters in the second image processing model according to the first depth estimation feature, the second depth estimation feature, the predicted depth image and the label depth image to generate a lightweight image processing model, and performing depth image processing on the image based on the lightweight image processing model. By adopting the method, the accuracy of extracting the image depth features by the model can be improved.

Description

technical field [0001] The present application relates to the technical field of the Internet, and in particular to an image processing method, device, equipment, and computer-readable storage medium. Background technique [0002] With the rapid popularization of deep learning technology and the improvement of computing power, artificial intelligence has been applied to various fields. [0003] The depth estimation of the game scene is an important part of training the artificial intelligence (AI) of gunfight games. Through depth estimation, computers and other machines can obtain the distance between different objects in the environment where the game character is located and the distance between the game character. For different Depending on the distance, the game AI can take different actions, so it can avoid collisions with obstacles, which can improve the effect of the game AI. [0004] In the prior art, the small model used to estimate the depth of the game image, tha...

AI Technical Summary

Problems solved by technology

[0004] In the prior art, the small model used to estimate the depth of the game image, that is, the model with a small number of parameters, although the hardware requirements for the user terminal are relatively low, and it is more convenient to deploy on the user terminal, but due to the depth of the small model Shallow, the feature representation ability is limited, so the small model has the defect of low accuracy of the extracted depth features, that is, there will be a large error between the estimated depth image and the real depth image

Images