Scalable structure learned via context-independent recursive document decomposition

• Main Authors: ARYA ARUN, GOYAL MAYANK
• Format: Patent
• Language: Chinese; English
• Published: 15.04.2022
• Subjects: CALCULATING; COMPUTING; COUNTING; PHYSICS

A method is provided in which a document is converted into a bitmap image, and a processing method aggregates pixel values from the bitmap image into a set of row sum values and a set of column sum values. The bitmap image is a pixelated representation of the document. The method applies a local Fourier transform to the set of row sum values and the set of column sum values to generate a frequency representation of the set of row sum values and a set of frequency sum values. The method decomposes the bitmap image into a set of image portions based on at least one separation location identified in the set of frequency representations, and sends the set of image portions to the text recognition system. 提供了一种方法，其中文档被转换成位图图像，并且处理方法将来自位图图像的像素值集聚合成行总和值集和列总和值集。位图图像是文档的像素化表示。该方法对行总和值集和列总和值集应用局部傅立叶变换以生成行总和值集的频率表示和频率总和值集。该方法基于在频率表示集中识别的至少一个分离位置，将位图图像分解成图像部分集，并且将图像部分集发送到文本辨识系统。