Image feature coding and recognizing method based on integrated learning

Abstract

The invention discloses an image feature coding and recognizing method based on integrated learning and relates to the technical field of signal and information processing. The method comprises the steps that training samples are established, and the sampling probability of the samples is initialized; according to the sampling probability, training samples are selected, a feature coding function is trained through the samples obtained through sampling, and the training samples are classified; according to a classification result, the sampling probability of the training samples is updated in a self-adaptive mode; the sampling and training steps are executed in a circulatory mode until a training stop condition is met. The feature coding function obtained through training can map any image feature vector into a short hash sequence, the distance between hashes can be made to be matched with the sensing similarity between images to the maximum extent, and the method has the advantage that the calculation complexity is low. A test result shows that the image hashes generated according to the method can provide high recognition accuracy during image content recognition.

Description

technical field [0001] The invention relates to the technical field of signal and information processing, in particular to an image feature encoding and recognition method based on integrated learning. Background technique [0002] Content recognition is the core technology to solve image query, indexing and copyright management problems. The goal of image content recognition is to search for homologous images with the same perceptual content as the image to be queried in a large database or network. The premise of image content recognition is to describe the perceptual content of the image. To reduce the complexity of image content recognition, it is usually necessary to encode image features into short descriptors. By comparing the descriptors, the perceptual similarity of the images can be judged. In order to achieve accurate image content recognition, the original image and its homologous version (such as the version of the original image after lossy compression, filte...

AI Technical Summary

Problems solved by technology

However, existing studies have shown that the distance between the quantitative indexes of features is difficult to accurately reflect the similarity of image perception content, which affects the performance of image content recognition.

Images