Image processing device, image processing method and information processing device

Abstract

The invention provides an image processing device, an image processing method and an information processing device. The image processing device comprises a transformation unit, a binaryzation unit, a stability determining unit, and a stroke determining unit, wherein the transformation unit is used for transforming a to-be-processed image of an object which contains a preset color into a color space used for representing the preset color; the binaryzation unit is used for carrying out binaryzation on the image which is transformed into the color space on the basis of a plurality of binaryzation threshold values to obtain a plurality of binaryzation images; the stability determining unit is used for obtaining the stroke feature of the pixel on a corresponding pixel position on each binaryzation image in the plurality of binaryzation images, and determining the stability of the obtained stroke feature of the pixel in the plurality of binaryzation images; and the stroke determining unit keeps he pixel of which the stability of the stroke feature is greater than the stability threshold value as the stroke of the object.

Description

technical field [0001] The present disclosure generally relates to the field of image processing, and in particular, relates to an image processing device, an image processing method, and an image processing device for determining a desired object portion in an image. Background technique [0002] In the field of image processing, the results of processing or recognizing objects in images can have a very wide range of applications, such as object recognition, text analysis, and so on. [0003] Existing object recognition algorithms may be limited by the outer contours of objects. For example, there are a large number of seals in Chinese ancient books, and these seals have important research value and collection value for archaeologists, ancient book researchers, and collectors. It is hoped that the part of the seal can be determined from the ancient documents. In order to determine the part of the seal, generally, the algorithm in the prior art needs to make assumptions ab...

AI Technical Summary

Problems solved by technology

In order to determine the part of the stamp, generally, the algorithms of the prior art need to make assumptions about the shape of the external contour of the stamp, thus only circular, oval or rectangular stamps of predetermined shapes can be detected

In addition to the application of identifying seals from ancient texts, there are a large number of other applications for determining specific objects from images, such as determining the area containing the text part from the packaging of goods, and the part of the music score from printed matter etc. Existing algorithms for these applications may also be constrained by the overall outer contour of the object

Images