Image searching feedback method based on contents

Abstract

The invention relates to an image searching feedback method based on contents, which comprises selecting a training sample from an image data base, training the training sample by utilizing a support vector machine to obtain a feature classification model, and classifying an image into a visual category; determining the visual category of the image submitted by a user according to the feature classification model, searching an image similar to the image submitted by the user, and returning a searching result; selecting feedback images in the searching result, marking images to be positive andnegative feedback images respectively according to the relevance between the feedback images and the image submitted by the user, and feeding back the mark result to a searching system; judging the accurate category of the image submitted by the user according to text keywords of the feedback images, category labels and a mapping table, searching the image similar to the image submitted by the user in the accurate category, and returning a secondary searching result. The image searching feedback method based on contents can fast and accurately locate the category the image submitted by the user belongs to and improve secondary searching accuracy.

Description

technical field [0001] The invention belongs to the field of image retrieval and recognition, and more specifically, the invention relates to a content-based image retrieval feedback method. Background technique [0002] The retrieval accuracy of traditional content-based image retrieval technology is usually not ideal, and relevant feedback technology can make up for this shortcoming to a certain extent. People have done a lot of research work on feedback, and have achieved a lot. A typical example is the query point moving method proposed by Rui et al. This method modifies the feature vector of the image submitted by the user by imitating the Rachio formula in text retrieval, so that it moves in the direction expected by the user. The modified eigenvector is the weighted sum of the original eigenvector of the image submitted by the user, the eigenvector of the positive feedback image and the eigenvector of the negative feedback image, so that it is biased towards the eige...

AI Technical Summary

Problems solved by technology

[0003] However, the existing image retrieval feedback methods have the following problems: due to the existence of semantic gap, the traditional method of changing the image features submitted by users has little effect on improving the accuracy of secondary retrieval results; the role of feedback is only for the current retrieval, The accuracy is still very low when the same image is submitted for retrieval next time; the feedback method of machine learning introduces machine learning during retrieval, so it is difficult to guarantee real-time performance. At the same time, due to the small number of samples, the training effect is not obvious, which has little effect on improving retrieval accuracy. Big

Images