No abstract

To resist geometrical attacks, the video zero watermarking algorithm based on logpolar transform presented in this paper. In our method, an original image transformed in logpolar coordinate after transformation of 2D DWT and 3D DCT. In experiment, the proposed method was evaluated the performance of resistance against attacks such as noise attack, rotation attack, compression attack and frame attack. The experiment results show that this algorithm can effectively resist against geometric attacks, and it has high robustness to the noise, filtering, compression and other common attacks. The bit error rate of the proposed algorithm is less than 0.06 for all tested attacks.

No abstract

This paper proposes a zero-video watermark scheme based on 2D-DWT and pseudo 3D-DCT, otherwise Singular Value Decomposition and log-polar transform are applied. The method makes no changes to original images while embedding the owner information of images so as to achieve high transparency. The log-polar transform ensures that the method is robust to rotation operations. In order to achieve the high robustness and security, we use chaotic logistic mapping and spread spectrum (CLMSS) to spread the watermark information. we also use the visual cryptography (VC) scheme to split the secret image into two shares. In the scheme, we use spread spectrum to encode the watermark to a Code Division Multiple Access watermark and use dual transform and log-polar to generate the feature values. Then the visual cryptography scheme is applied to generate the secret image from the feature values and the watermark. In the extraction scheme, we use the secret image which is registered to certification authority and the feature values extracted from the examined image with visual cryptography scheme to generate the CDMA watermark, and then decrypt it to get the watermark information. The experimental are conducted to verify the robustness through a series of experiments.

This paper presents a new content-based image retrieval (CBIR) method based on image feature fusion. The deep features are extracted from object-centric and place-centric deep networks. The discrete cosine transform (DCT) solves the strong correlation of deep features and reduces dimensions. The shallow features are extracted from a Quantized Uniform Local Binary Pattern (ULBP), hue-saturation-value (HSV) histogram, and dual-tree complex wavelet transform (DTCWT). Singular value decomposition (SVD) is applied to reduce the dimensions of ULBP and DTCWT features. The experimental results tested on Corel datasets and the Oxford building dataset show that the proposed method based on shallow features fusion can significantly improve performance compared to using a single type of shallow feature. The proposed method based on deep features fusion can slightly improve performance compared to using a single type of deep feature. This paper also tests variable factors that affect image retrieval performance, such as using principal component analysis (PCA) instead of DCT. The DCT can be used for dimensional feature reduction without losing too much performance.