Image content recognizing method and recognition system

Abstract

The invention discloses a picture content recognition method and system. It includes: pre-storing basic image content instances and corresponding annotation information in the instance library. When identifying, the splitting module splits one or more basic image contents from the picture to be identified; the basic image contents split out are compared with the basic image content instance in the example library by the similarity comparison module Comparing to obtain the corresponding similarity; determining the basic image content instance with the highest similarity with each basic image content by the result output module, and outputting the annotation information corresponding to the basic image content instance as the picture content recognition result. By using the invention, the calculation overhead of the computer and the occupation of system resources can be reduced, and the types of identifiable content can be expanded.

Description

technical field [0001] The invention relates to computer information processing technology, in particular to a recognition system and a recognition method for automatically recognizing computer picture content. Background technique [0002] In the current computer Internet identity verification control system, picture verification code technology is often used to realize customer identity verification and use frequency control. figure 1 It is a schematic diagram of an interface that realizes authentication and login through a picture verification code. see figure 1 , the user not only needs to input the correct account number 101 and password 102, but also needs to correctly read the verification code characters 103 from the verification code picture 100 and input them into the system, so as to pass the verification and log in to the system. [0003] In the process of testing the verification code system, in order to evaluate the risk of the verification code system, a tes...

AI Technical Summary

Problems solved by technology

[0006] 1) The picture content recognition unit needs to integrate various statistical algorithms and artificial intelligence algorithms to realize, the calculation process is complicated, the calculation amount is very huge, and a large amount of computer CPU operation overhead and system resources are required

[0007] 2) Only specific text can be identified from the picture, and it is difficult to identify image information

For example, if the picture also includes some specific patterns, such as a pattern of a sheep, the corresponding verification code input may be the word "sheep". In this case, the existing recognition technology cannot recognize

Images