Paper laboratory sheet image preprocessing method and system

Abstract

The invention discloses a paper laboratory sheet image preprocessing method and system. The method comprises the following steps: constructing a laboratory sheet image preprocessing model comprising a background estimation sub-network, a shadow removal sub-network and a crease removal sub-network; carrying out feature learning on the paper laboratory sheet image by adopting a background estimation sub-network to obtain a global background color and spatial distribution features of background and non-background pixels so as to construct a shadow attention map; training the shadow removal sub-network according to the original paper laboratory sheet image, the global background color and the shadow attention map to obtain a shadow-free image; and constructing a crease attention graph by adopting the background estimation sub-network according to the shadow-free image, and training the crease removal sub-network according to the shadow-free image and the crease attention graph to obtain a shadow-free crease-free laboratory sheet image. The visual quality and readability of the image are improved, and it is ensured that the good effect and robustness are achieved for laboratory sheet images with different features.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to an image preprocessing method and system for a paper test sheet. Background technique [0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art. [0003] At present, paper test sheets are still the main carrier of hospital inspection reports, but they are not easy to save. With the popularity of mobile phones, people are more inclined to use mobile phone cameras to save paper documents, but there are often problems in doing so. For example, the geometric shape of the document may be distorted due to the angle of the camera, and shadows may be formed due to the occlusion of the light source. In particular, when the paper test sheet itself has creases, the problem of uneven light distribution on the page is particularly prominent . The above-mentioned problems will lead to p...

AI Technical Summary

Problems solved by technology

Using the method of deep learning, two convolutional neural networks are respectively proposed to solve the document image geometry and lighting distortion, but the lighting correction network is not effective in processing image shadows and creases.

Images