Topic segmentation method and device based on target area fusion, equipment and medium

Abstract

The embodiment of the invention relates to the technical field of topic detection, and discloses a topic segmentation method and device based on target area fusion, equipment and a medium. The methodcomprises the steps of obtaining a target picture, and obtaining a topic region mask and text line information of the target picture; determining a target text line mask according to the target topicarea mask, and calculating a first intersection area of the target text line mask and the target topic area; determining a first proportion of the target text line in the target topic area according to the first intersection area and the area of the target text line; and when the first proportion is greater than or equal to a first preset threshold, expanding the target topic area through the target text line mask to obtain an expanded target topic area. By implementing the method of the invention, topic segmentation can be supplemented through character recognition, so that the fused boundarycannot cut the text line, and meanwhile, the result of the character information is combined into the topic detection area, so that the topic detection result is more complete.

Description

technical field [0001] The present invention relates to the technical field of topic detection, in particular to a topic segmentation method, device, electronic equipment and storage medium based on fusion of target areas. Background technique [0002] At present, there is an end-to-end topic segmentation algorithm based on deep learning. This algorithm can divide the topic areas of documents such as test papers and exercise books. However, due to the black-box characteristics of deep learning, the results obtained are not ideal. There are the following problems : [0003] 1. The divided area is not accurate enough, and the text is often cut, resulting in the lack of information and affecting subsequent use. [0004] 2. It is not sensitive to the distinction of text lines, cannot obtain specific texts, and cannot fully mine the information contained therein. Contents of the invention [0005] In view of the above defects, the embodiment of the present invention discloses...

AI Technical Summary

Problems solved by technology

[0002] At present, there is an end-to-end topic segmentation algorithm based on deep learning. This algorithm can divide the topic areas of documents such as test papers and exercise books. However, due to the black-box characteristics of deep learning, the results obtained are not ideal. There are the following problems :

[0003] 1. The divided area is not accurate enough, and the text is often cut, resulting in the lack of information and affecting subsequent use

[0004] 2. It is not sensitive to the distinction of text lines, cannot get specific texts, and cannot fully mine the information contained in them

Images