Search

In the area of full-reference (FR) objective ‘image quality assessment’ (IQA), a huge amount of improvement has been done in the last few years that can calculate image quality, consistently. Contrarily, query-based image/video databases and search engines retrieve related data using ‘ranking and indexing’ depending on stored content. It is also to be noted that both techniques use feature extraction to achieve their goal. The efficiency of seven selected FR-IQA schemes is described in this article for the retrieval of an image signal. Extensive tests are done on two freely accessible databases. The comparison results express that mean-structural-similarity-index-measure (MSSIM), and feature-similarity-index-measure (FSIM) offer superior outcome than other IQA models. Our assessment outcome and the related thought will be very much useful for the researchers to understand the latest application areas of the IQA model in the area of image searching and retrieval. more...

This paper proposes a data-hiding scheme for quality access control of digital images using quantization index modulation (QIM) and its hardware implementation. To achieve the goal, an encoded binary message is embedded over N-mutually orthogonal signal points using the QIM technique but without complete self-noise suppression. It is well known that due to the insertion of external information, there will be degradation in the visual quality of the host image. This feature may be used in access control through the reversible process. At the decoder side, watermark bits are extracted using minimum distance decoding. Self-noise is then suppressed by the authorized user to provide a better quality of the image. Moreover, for real-time implementation, field-programmable-gate-array (FPGA) based hardware architecture is also proposed. The scheme is tested over a large number of benchmark images, and the experimental results are compared with the related scheme and found to be superior. It is also seen that (a) in real-time processing, the scheme saves 87.86% power than the related implementation found in the literature, (b) a very high throughput of 119.54 Megabyte/s and 119.048 Mbps are achieved for encoder and pipelined decoder at the maximum operating frequency of 120.013 MHz and 120.01 MHz, respectively for the processing of (512 × 512) sized images. more...

In this paper, Discrete Cosine Transform (DCT) based blind video watermarking algorithm is proposed, which is perceptually invisible and robust against rotation and collusion attacks. To make the scheme resistant against rotation, watermark is embedded within the square blocks, placed on the middle position of every luminance channel. Then Zernike moments of those square blocks are calculated. The rotation invariance property of the Complex Zernike moments is exploited to predict the rotation angle of the video at the time of extraction of watermark bits. To make the scheme robust against collusion, design of the scheme is done in such a way that the embedding blocks will vary for the successive frames of the video. A Pseudo Random Number (PRN) generator and a permutation vector are used to achieve the goal. The experimental results show that the scheme is robust against conventional video attacks, rotation attack and collusion attacks. more...

Nowadays, the quantization index modulation (QIM) principle is popular in digital watermarking due to its considerable performance advantages over spread-spectrum and low-bit(s) modulation. In a QIM-based data-hiding scheme, it is a challenging task to embed multiple bits of information into the host signal. This work proposes a new model of QIM, i.e., the M-ary amplitude modulation principle for multibit watermarking. The watermark embedding process may be divided into two phases. In the first phase, a binary watermark image is spatially dispersed using a sequence of numbers generated by a secret key. In the second phase, the host image is decomposed by lifting, and the encoded watermark bits are embedded into the high–low (HL) and low–high (LH) subbands of DWT-coefficients using M-ary amplitude modulation. The results of the simulation show that the robustness increases, at the cost of increased decoding complexity, for a high M-value. Furthermore, this investigation has shown that the decoding complexity of higher M-values can be overcome at moderate N-values, while the robustness performance is maintained at satisfactory level. more...

Transmission of block-coded images through error prone radio mobile channel often results in lost blocks. To this aim, this paper proposes a protection mechanism for discrete cosine transform based compressed image(s) through error concealment without affecting the compatibility of standard JPEG coding. The goal is achieved by embedding a low-resolution version (image-digest) of an image into itself during encoding using M-ary quantization index modulation. Halftoning technique is used to obtain the image digest from the host image. At the decoder, this embedded watermark is extracted first and is then used as a reference one to conceal the transmission losses in the received image. Mathematical analysis as well as simulation results show that the use of M-ary modulation offers significant performance improvement than traditional quantization index modulation based methods in large number of lost block (packet) scenario. Relative compensation for visual quality improvement is also tested in Rayleigh fading wireless channel and is compared with the few other existing methods. more...

This paper proposes a tamper detection and correction technique using semi-fragile data hiding that aims to achieve high perceptual quality of images at the user-end even after malicious modifications. A binary signature and an image digest are embedded by modulating integer wavelet coefficients using dither modulation based quantization index modulation. Half-toning technique is used to obtain image digest from the low-resolution version of the host image itself. Decoder extracts the binary signature from the watermarked image for tamper detection, while the extracted image digest is used to correct the tamper region. Unlike previously proposed techniques, this novel approach distinguishes malicious changes from various common image processing operations more efficiently and also correct tapered regions effectively. Experimental results show that the proposed technique provides a superior performance in terms of probability of miss and false alarm as well as in tamper correction, compared to several existing semi-fragile watermarking techniques. more...

This paper proposes a joint data-hiding and data modulation scheme to serve the purpose of quality access control of image(s) using quantization index modulation (QIM). The combined effect of external information embedding and data modulation cause visual degradation and may be used in access control through reversible process. The degree of deterioration depends on the amount of external data insertion, amount of data modulation as well as step size used in QIM. A weight function (q) and the quantization step size (@D"b) of JPEG 2000 are used for defining the step size of QIM. Lifting based discrete wavelet transform (DWT), instead of conventional DWT, is used to decompose the original image in order to achieve advantages, namely low loss in image quality due to QIM, better watermark decoding reliability and high embedding capacity for a given embedding distortion. At the decoder, data are demodulated first and watermark bits are then extracted using minimum distance decoding. Extracted watermark is used to suppress self-noise (SN) that provides better quality of image. Simulation results have shown that an improvement in peak-signal-to-noise-ratio (PSNR), mean structural-similarity-index-measure (MSSIM) and Watson distance by an amount of about 30%, 12% and 77%, respectively, are gained by authorized user when 50% coefficients of high-high (HH) coefficients are modulated. more...

This paper proposes a region based digital watermarking scheme to serve the purpose of copyright protection and integrity verification of image database using discrete cosine transform (DCT). The objective is achieved by embedding watermark into visually significant regions of the image(s) using quantization index modulation (QIM). The main contributions of this work are, (1) the use of quad tree decomposition for image segmentation to find visually significant regions, and (2) the use of image database as a single entity for watermarking, instead of watermarking on individual image, so that integrity of the database is checked. Simulation results validate the above claims. Simulation results also show that the proposed scheme is robust and secured against a wide range of attacks available in StirMark 4.0 package. more...

This paper proposes a scheme for quality access control of discrete cosine transform (DCT) based compressed color image(s) using data hiding. The goal of access control is achieved by transforming the original image into the luminance, chrominance-blue and chrominance-red ( YC b C r ) color space followed by modulating the coefficients of luminance (Y) and/or chrominance-red ( C r ) channels. The amount of modulation and the information about the specific channels to be modulated are contained in a key K, which is supplied by the owner/service provider. This key and the original image are used to create a content dependent key. The content dependent key is then embedded into the chrominance-blue ( C b ) component as watermark using spread spectrum modulation. This watermark information is used for demodulation that leads to the access control through the restoration of the relative quality of the images. Simulation results show that the proposed scheme provides an effective quality access control for the compressed color image without decreasing the compressibility of the standard JPEG coding scheme. The performance of the proposed method is also compared with the few other methods. more...

Transmission of digital data through radio mobile channel experiences fading effect that leads to severe degradation in quality. Error concealment is a postprocessing technique to improve this quality degradation. One such algorithm is proposed here using integer wavelet, data hiding and region-of-interest (ROI) coding functionality. Threshold based image segmentation and morphological operations are used together to find ROI. Halftoning technique is then used to obtain image digest from this ROI. To restore the human interested portion of an image, the information of ROI is embedded as image digest into the region-of-background (ROB) using data hiding based on quantization index modulation (QIM). Simulation results show that the scheme provides significant performance improvement with various lost blocks in ROI without affecting the compatibility of the standard JPEG 2000 coding scheme. Performance of the proposed scheme is also tested using simulations on Rayleigh-fading wireless channel and is compared with the few other error concealment methods. more...