Research On Digital Watermarking Algorithm Of Remote Sensing Image With High Radiometric Resolution

High radiometric resolution (≥10bit) digital remote sensing image obtained by aeronautics and astronautics technology has gradually been the main data acquisition source of spatial geographic information. Compared with common remote sensing image, high radiometric resolution remote sensing image can distinguish more precise details of various surface features, and lower the effect of mixed-pixel for quantitative analysis, improve the accuracy of remote sensing analysis, so it has broad application prospects. Therefore, the protection of its data security seems especially important. As a research focus of information hiding field, digital watermarking technique can provide effective security support for high radiometric resolution remote sensing image. This paper firstly analyzes the characteristics of high radiometric resolution remote sensing image, summaries the present watermarking algorithms, and then discusses the digital watermarking technology requirements of high radiometric resolution remote sensing image, based on these, researching the digital watermark algorithms that applicable to its own characteristics. The main works are as following:1. Detailed introducing of the digital watermarking technique, including principles, characteristics, classification, applications and the current research condition; analyzes the characteristics of high radiometric resolution remote sensing image and its differences with the general remote sensing image, then discusses the digital watermarking algorithm technical requirements of high radiometric resolution remote sensing image and the quality evaluation methods.2. A reversible visible watermarking algorithm for high radiometric resolution remote sensing images'notification is realized. To keep the important visual characteristics of image coverage area by using a bit-shift strategy to push down high bits,and uses Huffman coding to compress the watermark information and the low bits of carrier image, the original image can be non-destructive reconstructed without knowing any information about the watermark on the receiving side. This algorithm has a better visual impression, the watermark can be embedded into any position of the carrier image without brightness transition, the coverage area is translucent-like, and carrier image content can still be roughly identified.3. The proposition and realization of an image watermarking scheme based on feature points and Dual-Tree Complex Wavelet Transform. It firstly calculates three Complex Wavelet Transform of the origin image, then uses F?rstner operator extracting the feature points in low-frequency sub-band, according to the characteristics of complex wavelet transform coefficients, we can implement the watermark modulation embedding by modifying the corresponding high-frequency coefficients, when the watermark will be extracted, the original carrier image can be near-lossless recovered at the same time. Algorithm has both robustness and invisibility, can effectively resist cropping, translation, rotation and other geometric attacks, and also has some robustness to linear stretching and low-pass filtering attacks.4. Improved the lossless covert communication method based on integer linear transform, designs a reversible digital watermarking algorithm for local interest target authorization exploitation in remote sensing image. First, extracting the local interest target in remote sensing image as the watermark to be hidden, deforming the target on the original image, and then using reversible integer linear transform to achieve the watermark modulation embedding, the authorized users can fully lossless reconstruct the original remote sensing image. The algorithm is blind, with a large embedding capacity, and the watermarked image has a high peak signal to noise ratio, and can effectively resist cropping, salt-pepper noise and other attacks.5. Aiming at resisting the linear stretching attack which usually been done to the high radiometric resolution digital remote sensing image, presents a reversible watermarking algorithm based on integer DCT transform. According to the high correlation between adjacent pixels in the same image sub-block, the watermark can be embedded by using a high frequency coefficient modification technique in the integer DCT domain, the corresponding embedding capacity increase scheme and distortion control scheme are also proposed. During the information extraction process, the original image can be perfectly restored with the embedded information accurately extracted at the same time. The results of experiment proved that this algorithm is invisible, can efficiently against linear stretching attack, and has better robustness to cropping, rotation, noise or filter attacks.