An image processing and character segmentation method based on a picture verification code

Abstract

The invention provides an image processing and character segmentation method based on a picture verification code. The method comprises the following steps: S1, carrying out binarization processing onthe picture verification code by using a low-pixel gray scale median method; S2, repairing the character edge in the binarized image by using a four-connected domain or eight-connected domain algorithm; S3, de-noising the restored image by using a template statistics selection filter; S4, calculating and correcting a character segmentation line based on a reference line vertical segmentation method, and segmenting the denoised image into R character samples, R being the number of characters in the image; and S5, calculating the cutting position of the redundant part of each character sample,and cutting the character according to the position to obtain the character sample of which the redundant part is cut off. The method is simple in calculation, the image processing and character segmentation precision is higher, the applicability is better, the character samples processed through the method are more convenient for subsequent machine recognition, and the machine recognition accuracy is improved.

Description

technical field [0001] The invention relates to an efficient image processing and character segmentation method for picture verification codes, and belongs to the technical field of picture verification codes. Background technique [0002] Image verification code technology is a part of modern network security technology, which is used to distinguish whether the access object is a human or a robot, so as to prevent purposeful access with large traffic in the program. General verification codes can prevent some simple mechanical access, but in today's era when machine recognition is mature, it is no longer a problem to recognize pictures of simple numbers and letters. Many scholars at home and abroad have conducted research on the identification of verification codes. The purpose of trying to recognize verification codes by machines is not to crack verification codes but to invent more effective verification codes that are difficult for machines to recognize, thereby maintain...

AI Technical Summary

Problems solved by technology

Research at this stage shows that character segmentation is the most difficult point in captcha processing. At present, there is no universal and perfect algorithm to solve the problem of segmenting complex cohesive characters. Even some captchas require a certain amount of time for human eyes to distinguish, which hinders the machine. The main obstacle to identifying the verification code has also become the basic guarantee that the verification code can still become a means of distinguishing between man and machine.

Foreign scholars pointed out that when the verification code of a group of samples can be effectively recognized by the machine (completely recognized correctly) more than 10%, the verification code (system) is no longer safe

Images