Image processing method and device, storage medium and electronic equipment

Abstract

The embodiment of the invention discloses an image processing method and device, a storage medium and electronic equipment. The method comprises the following steps: acquiring a to-be-processed image, and inputting the to-be-processed image into a shadow prediction model to obtain a shadow image of the to-be-processed image; and inputting a to-be-processed image into the albedo prediction model to obtain initial albedo distribution of the to-be-processed image, inputting the to-be-processed image into the albedo edge prediction model to obtain an albedo edge image of the to-be-processed image, and fusing the initial albedo distribution and the albedo edge image to obtain an albedo image of the to-be-processed image. According to the method, a prediction model is independently designed for each intrinsic component, so that the characteristics of each component are independently coded, interference among components when a single prediction model is used for prediction is avoided, and the intrinsic decomposition quality is improved; in addition, each intrinsic component corresponds to one prediction model, the method can be suitable for more complex scenes, and the intrinsic decomposition quality under complex illumination and texture conditions is improved.

Description

technical field [0001] Embodiments of the present invention relate to image processing technologies, and in particular to an image processing method, device, storage medium, and electronic equipment. Background technique [0002] In the field of image processing, the physical composition of images is crucial for computer vision and graphics applications, and the method of extracting image processing is called intrinsic image decomposition. [0003] figure 1 It is a flowchart of the existing intrinsic image decomposition method. There are two existing intrinsic image decomposition schemes: 1. Use one network to predict the albedo image and shadow image at the same time; 2. Use only one network to predict the albedo from the input image The albedo image is then calculated by dividing the albedo by the albedo image to obtain the shadow image. [0004] In the above two schemes, in scheme one, since the albedo image and shadow image have different image characteristics, using t...

AI Technical Summary

Problems solved by technology

[0005] The present invention provides an image processing method, device, storage medium and electronic equipment to solve the problems in the prior art that the same prediction network simultaneously predicts two eigencomponents to form mutual interference and cannot be applied to complex illumination. scene problem, improving the predictive performance of each eigencomponent

Images