Form recognition method and device, computer storage medium
A recognition method and table technology, applied in the information field, can solve problems such as the failure to restore the table, technology cannot successfully identify and restore the table, etc., and achieve the effect of improving the recognition success rate and integrity
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
example 1
[0144] OCR table detection technology, while recognizing the table content (for example, text), also detects the table structure, which greatly reduces the post-processing cost of OCR table recognition, and realizes a complete set of end-to-end recognition and typesetting. This technology is aimed at Table text recognition has many application scenarios, such as receipts, bills, and various forms of text data.
[0145] This technical solution is a table recognition method based on a deep learning model. On the basis of the current OCR text detection and recognition, table structure detection is added. Use X, Y, W, H to represent the cell, where (X, Y) is the coordinates of the center of the cell, W, H are the width and height, and the features of the last layer of the deep neural network are predicted by the points inside the cell X, Y, W, H of the cell. Subsequent processing performs maximum interference suppression algorithm NMS processing in the predicted multiple cell fra...
example 2
[0148]Such as Figure 5 As shown, this example provides a table recognition method, including:
[0149] Input the image with recognition to the deep neural network;
[0150] The deep neural network outputs text detection results, text recognition results and form detection results;
[0151] Integrating text detection results, text recognition results and table detection results, a table with text information will be output.
[0152] Figure 6 for input to Figure 5 The original form of the deep neural network shown; Figure 7A For text detection results, obviously in Figure 7A In the text, the text is represented by a chamfered rectangle. Figure 7B For the text recognition results, use OCR technology to convert the table content in the image into text information, and combine the text detection results and text recognition results to know which of the recognized text information belongs to the same cell and which text information belongs to different cells. of cells. ...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More