Image hash processing method based on adjacent gradient and structural features

Abstract

The invention discloses an image hash processing method based on adjacent gradients and structural features, and the method comprises the steps: reading an image in an image library, and carrying out the preprocessing of the image; three components of the preprocessed image being extracted, and obtaining statistical characteristics of the image by using a near gradient and binarization quantization compression method; converting the preprocessed image into a color space, extracting a brightness component of the image, and converting the brightness component into a three-dimensional image; extracting structural features of the image according to the brightness component image and the three-dimensional image; and combining the statistical features with the structural features to obtain an intermediate hash, and performing position scrambling on the intermediate hash by using a random generator to obtain a final hash sequence. According to the invention, the method has robustness for processing most of images with kept contents, blocks the images, has no robustness for large-angle rotation, and has a very low collision rate; for image authentication, features in an image library are utilized to form image hash, and high safety performance is achieved.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to an image hash processing method based on adjacent gradients and structural features. Background technique [0002] In recent years, the content security of digital media has been widely concerned by people. With the rapid improvement of technology and network, as well as smart multimedia and the popularization of various image editing software, more and more images and videos are generated by users and uploaded to the Internet community. People can easily obtain a large amount of image information from it, and use various software such as photoshop, time magic hand, etc. to perform simple operations on images, such as adding text to images, changing image brightness, contrast, and synthesizing new images. Therefore, an image may produce many copies, so how to distinguish these images is very important. As a result, the technique of image hashing came into being to dete...

AI Technical Summary

Problems solved by technology

[0003] There are many outstanding algorithms made by researchers in the field of image hashing. For example, Lei et al. combined Radon transform and discrete Fourier transform (discrete Fourier transform, DFT) to construct a hash, by extracting the invariant image after transformation The features and the DFT coefficients of one-dimensional DFT transformation are quantized to form a hash. Qin et al. [2] considered increasing the anti-rotation ability of the algorithm while maintaining the distinction, so they extracted the largest inscribed circle of the image to reconstruct the image block, and used DFT is transferred to the frequency domain, and non-uniform sampling is used to extract robust features from the magnitude matrix of Fourier coefficients to form a hash, etc. There are problems such as weak robustness to gamma correction and Gaussian noise, and affecting image classification performance.

Images