Image semantic automatic annotation and retrieval method and system

Abstract

The invention belongs to the field of image annotation and retrieval, and discloses an image semantic automatic annotation and retrieval method and system. The method comprises the following steps: collecting a training set image, segmenting the training set image, and extracting feature elements; collecting images of the detection set, segmenting the images of the detection set, and extracting feature elements; carrying out image classification on the training set image and the detection set image according to the similarity of the characteristic elements of the training set image and the detection set image; performing semantic annotation on images of the detection image set; and inputting related keywords into the search interface to search the image. According to the method, the imagesare classified through feature element extraction of the images of the intelligent system, semantic definition is conducted on the image set, the working efficiency is improved, the workload of workers is reduced, and the detection effect is improved.

Description

technical field [0001] The invention belongs to the field of image annotation and retrieval, in particular to an image semantic automatic annotation and retrieval method and system. Background technique [0002] At present, the existing technologies commonly used in the industry are as follows: with the advent of the information society and the popularization of computer applications, people are increasingly exposed to a large amount of information, among which multimedia information is the most widely accessible information resource , it exists in various forms such as text, image, sound and video, and grows at an explosive speed with the advancement of technology. Especially in recent years, the application and development of the Internet has further promoted the rapid growth of the data volume of multimedia information. In the face of massive data, people are often at a loss. The rapid growth of information makes people's demand for multimedia information retrieval is in...

AI Technical Summary

Problems solved by technology

[0004] (1) In the prior art, images are manually classified and semantically defined, which increases the workload of staff and reduces work efficiency

[0005] (2) The running speed of traditional algorithm image classification is slow, and the classification accuracy rate is low

[0006] (3) In the process of extracting relevant features from the segmented image using the traditional algorithm in the prior art, the color and texture information of the image cannot be effectively extracted, which reflects the spatial correlation of each feature and reduces the retrieval effect

[0007] (4) In the process of using the traditional algorithm to calculate the similarity between the two, it cannot tolerate a small amount of modification to the picture, and cannot avoid the influence of gamma correction or color histogram adjustment

Images