A method and device for image recognition

Abstract

The present invention provides an image recognition method and device, the method comprising: establishing a data space in advance, one coordinate axis in the data space is a color axis; acquiring the color information of each pixel in the image to be recognized; according to the The color information of each pixel of the image to be recognized, loading each pixel of the image to be recognized to the corresponding position of the color axis in the data space, so as to form at least one layer on the color axis ; Identify each layer. The invention provides an image recognition method and device, which can improve the image recognition rate.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to an image recognition method and device. Background technique [0002] Image recognition refers to the technology of using computers to process, analyze and understand images to identify targets and objects in various patterns. OCR (Optical Character Recognition, Optical Character Recognition) technology is a kind of image recognition technology, which is widely used in image content recognition. [0003] In the prior art, the image is mainly recognized through the OCR technology. Obtain the image of the bill, draw the outline of the characters in the image of the bill by optical means, and then compare it with the font template in the standard font library to find a matching font template. [0004] It can be seen from the above description that the recognition methods in the prior art always try to analyze and recognize the geometric structure of characters through hig...

AI Technical Summary

Problems solved by technology

[0004] It can be seen from the above description that the recognition methods in the prior art always try to analyze and recognize the geometric structure of characters through high-contrast image content. When the colors in the image are complex or the colors between the geometric structures are close, the character low recognition rate

Taking bank cash checks as an example, not only the background pattern of the bill itself is complex and the colors are changeable, but also the machine-printed numbers are mostly in color, and the colors of the original icons, fonts, seals, and handwritten texts are also different. If it is difficult to get the correct recognition result by using traditional OCR recognition technology

In short, the recognition rate of the image content is related to the content complexity of the image itself. The higher the color contrast between the background and the text, the fewer the colors contained, the higher the recognition rate, and vice versa.

Images