Table picture recognition method and broadcasting method, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a table picture recognition method and broadcasting method, electronic equipment and a computer readable storage medium, and relates to the technical field of image processing. The table picture recognition method comprises the following steps: recognizing a table in a picture to determine type information of a plurality of cells contained in the table, wherein the type information of the cells at least comprises a value type and a keyword type; and based on the position arrangement of the cells of each type in the table, determining a mapping relationship between the cells of each value type and the cells of the associated keyword type. The mapping relationship between the table cells of each value type and the table cells of the associated keyword type is determined based on the association relationship among the plurality of table cells, so that the semantic information of the table can be expressed more accurately.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to a method for identifying a table picture, a method for broadcasting, an electronic device and a computer-readable storage medium. Background technique [0002] Currently, various text recognition technologies, such as Optical Character Recognition (OCR), have been widely used to recognize text in pictures. For pictures containing tables, the traditional text recognition technology either only recognizes the text part and ignores the table part, or although it can recognize the cells in the table, it is only at the image processing level from the picture as it is. Extract each cell and place the recognized text into the corresponding cell. [0003] It can be seen that the above technology only recognizes the text in each cell itself, and cannot obtain the semantic information of the text in the cell, nor can it obtain the logical relationship of the text between ...

AI Technical Summary

Problems solved by technology

For pictures containing tables, the traditional text recognition technology either only recognizes the text part and ignores the table part, or although it can recognize the cells in the table, it only recognizes the original text from the picture at the image processing level. Extract each cell and put the recognized text into the corresponding cell

[0003] It can be seen that the above technology only recognizes the text in each cell itself, and cannot obtain the semantic information of the text in the cell, let alone the logical relationship between the text in different cells

Therefore, when the text in the recognized cell is further processed (such as analysis or voice reproduction), the processing result consistent with the semantic information expected to be expressed in the table cannot be obtained

For example, when providing audio reading services for the hearing-impaired, due to the inability to accurately identify the logical relationship between cells, the reproduced voice may not conform to the semantic information that the table hopes to express.

Images