Science and technology literature image processing system based on deep learning

Abstract

A science and technology literature image processing system based on deep learning comprises: a data uploading and processing module used for uploading literature and image data; the data storage module that is used for storing the data processed by the data uploading and processing module; the deep learning algorithm module that is used for processing the data by utilizing a deep learning model;and the display and evaluation module that is used for beautifying and displaying the obtained image. Corresponding modules are constructed by adopting a deep learning technology, an image processingsystem is realized, the problem that scientific and technological literatures are difficult to find and extract by researchers due to too many scientific and technological literatures is solved, batchuploading of the scientific and technological literatures and extraction of all images in the literatures are supported, and analysis and understanding of the researchers are facilitated; color suggestions are provided for researchers to draw literature images, and the researchers are helped to draw images with rich and reasonable colors; the difficulty that image features need to be searched forthrough manual design in the past is solved through a deep learning algorithm, and literature images can be rapidly analyzed.

Description

technical field [0001] The present invention relates to deep learning and image processing, in particular to an image processing system for scientific and technological documents based on deep learning. Background technique [0002] With the popularization of the concept of artificial intelligence and the development of computer technology, deep learning technology has ushered in a breakthrough. Compared with traditional machine learning algorithms, deep learning technology has two advantages: First, as the amount of data continues to increase, it is difficult for traditional machine learning algorithms to improve performance, while deep learning technology can continuously improve its performance with the increase in data size Second, most traditional machine learning algorithms need to manually extract features, and researchers need to design different feature extractors for different problems, while deep learning technology can directly extract features from data, which g...

AI Technical Summary

Problems solved by technology

[0005] In order to overcome the deficiencies of the existing technology, the purpose of the present invention is to propose a deep learning-based scientific and technological document image processing system for the current situation that scientific researchers need to manually obtain a large number of document images, and the color usage is not clear when drawing. Realize the purpose of obtaining document images in batches and beautifying them

Images