Method and system for image classification, and method and system for image retrieval

Abstract

The invention provides a method and a system for image classification, and a method and a system for image retrieval. The image classification method includes dividing an image into a plurality of sub-regions; extracting a focus sub-region among the sub-regions; carrying out color semantic mapping on the focus sub-region and obtaining semantic characteristics; and carrying out image classification by utilizing image classification characteristics which include the semantic characteristics. According to the method and system for imaging classification and the method and the system for image retrieval, accuracy and processing efficiency of image classification are improved, complexity of classification processing is reduced, and image classification correlation is increased, and meanwhile accuracy and processing efficiency of image retrieval are enhanced, and usage experience of a user for image retrieval is improved.

Description

technical field [0001] The present application relates to the technical field of electronic image data processing, in particular to an image classification method, an image classification system, an image retrieval method, and an image retrieval system. Background technique [0002] With the rapid development of modern computer technology and Internet network, huge and diverse image information is constantly appearing, and images are used more and more widely in all walks of life, which further promotes the further development of image information resource management research. Now images are widely distributed on the Internet, lack of effective organization, it is difficult to achieve the purpose of resource sharing. Therefore, how to use massive image information and find out the required information is a major challenge to image information query technology. [0003] Since the early 1990s, the technology of retrieving images based on image content such as color, shape, te...

AI Technical Summary

Problems solved by technology

[0011] 1. Insufficient accuracy

[0012] The existing technology relies on the underlying digital features of the image in processing image features, and the pure underlying digital features cannot completely describe all the information of the image, so it will cause a large problem of insufficient accuracy in the classification results

For example, a photo containing a red profile horse and a photo containing a white front horse have a huge difference in their underlying features (color, shape, and texture). Simply relying on the underlying digital features will cause The classification of the two images is not accurate, that is, they may not both be classified into the image of "horse"

[0013] 2. High implementation complexity and parameter dependence

[0014] For the classification of images, the main amount of computation exists in the feature extraction and the construction of the classifier, and the implementation complexity is very high; moreover, the above-mentioned prior art needs to set more parameters, which has excessive dependence on the parameters, and the calculation time depends on all images

[0015] 3. The combination of the subjective meaning of the image is not high

[0016] Existing technologies that simply rely on the underlying digital features or text matching of keywords will affect the relevance of images. For example, "computer" and "computer" are equivalent concepts, but if these two words are used as two very different pictures image description, then the two images may not be assigned to the same category

Images