Digital watermark system based on copyright protection of Internet of things
A technology of the Internet of Things and watermarking, which is applied in biological neural network models, program/content distribution protection, image data processing, etc., and can solve problems such as poor robustness against attacks and inaccurate positioning of tampering
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment Construction
[0006] 1.1 RBF neural network
[0007] Radial Basis Function (RBF) neural network was proposed by J.Moody and C.Darken in the late 1980s. The radial basis function neural network consists of three layers, namely the input layer, the radial base layer, and the output layer. Radial Basis Neural Network is a typical feed-forward neural network. It has the advantages of fast training speed, the ability to converge to the global optimum, and the ability to approximate any continuous function with any precision, and has high application value.
[0008] 1.2 Watermark Quantization
[0009] 1.2.1 Quantification object
[0010] The classic quantization algorithm takes the coefficients in the time domain or frequency domain to quantize directly, and the signal-to-noise ratio is low. In this paper, the difference between the sum of the wavelet parity coefficients is selected as the quantization object. It is assumed that the wavelet coefficients are Y={y(i); 0≤i≤l}.
[0011] (1) Find ...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 