Model training and table recognition method and device

Abstract

The invention discloses a model training and table recognition method and device, and the method comprises the steps: determining a plurality of images containing a table as training samples, determining the mark of each training sample according to the structure and position of the table in the training sample, inputting the training sample into a feature extraction layer of a recognition model, and carrying out the recognition of the table. Determining an image feature pyramid corresponding to the training sample, for each feature map in the image feature pyramid, determining a reconstruction code corresponding to the feature map, performing up-sampling on the reconstruction code corresponding to the feature map, fusing the reconstruction code with other feature maps with the size larger than that of the feature map, and taking a fusion result corresponding to each feature map as input; and inputting a recognition layer of the recognition model to obtain a recognition result of the training sample. According to the method, fusion is carried out based on the feature maps of different sizes, the recognition result of the training sample is determined, the obtained image features are more comprehensive, abundant information can be obtained when the collected image is recognized, and the efficiency is high.

Description

technical field [0001] This specification relates to the field of computer technology, in particular to a method and device for model training and form recognition. Background technique [0002] Tables are widely used as an effective data organization and presentation method, and have also become common page objects in various documents. With the explosive growth of the number of documents, how to efficiently find tables and obtain content and structure information from documents, that is, table recognition, has become an urgent problem to be solved. [0003] In the prior art, a common form recognition method is implemented based on binarization. Specifically, an image containing a table may be acquired first, and binarization processing is performed on the image. Then, the binarized image can be respectively input into the cyclic neural network for row segmentation and the cyclic neural network for column segmentation to obtain the row segmentation result and column segme...

AI Technical Summary

Problems solved by technology

[0004] However, the existing technology can recognize tables in scanned images, but when recognizing images collected by image sensors, the tables in the collected images usually contain distortions, occlusions, etc., making the recognition effect relatively poor. Difference

Images