Reza Moradi Rad

A Unified Data Embedding and Scrambling Method

Conventionally, data embedding techniques aim at maintaining high-output image quality so that the difference between the original and the embedded images is imperceptible to the naked eye. Recently, as a new trend, some researchers exploited reversible data embedding techniques to deliberately degrade image quality to a desirable level of distortion. In this paper, a unified data embedding-scrambling technique called UES is proposed to achieve two objectives simultaneously, namely, high payload and adaptive scalable quality degradation. First, a pixel intensity value prediction method called checkerboard-based prediction is proposed to accurately predict 75% of the pixels in the image based on the information obtained from 25% of the image. Then, the locations of the predicted pixels are vacated to embed information while degrading the image quality. Given a desirable quality (quantified in SSIM) for the output image, UES guides the embedding-scrambling algorithm to handle the exact number of pixels, i.e., the perceptual quality of the embedded-scrambled image can be controlled. In addition, the prediction errors are stored at a predetermined precision using the structure side information to perfectly reconstruct or approximate the original image. In particular, given a desirable SSIM value, the precision of the stored prediction errors can be adjusted to control the perceptual quality of the reconstructed image. Experimental results confirmed that UES is able to perfectly reconstruct or approximate the original image with SSIM value after completely degrading its perceptual quality while embedding at 7.001 bpp on average.

Reversible data hiding by adaptive group modification on histogram of prediction errors

In this work, the conventional histogram shifting (HS) based reversible data hiding (RDH) methods are first analyzed and discussed. Then, a novel HS based RDH method is put forward by using the proposed Adaptive Group Modification (AGM) on the histogram of prediction errors. Specifically, in the proposed AGM method, multiple bins are vacated based on their magnitudes and frequencies of occurrences by employing an adaptive strategy. The design goals are to maximize hiding elements while minimizing shifting and modification elements to maintain image high quality by giving priority to the histogram bins utilized for hiding. Furthermore, instead of hiding only one bit at a time, the payload is decomposed into segments and each segment is hidden by modifying a triplet of prediction errors to suppress distortion. Experimental results show that the proposed AGM technique outperforms the current state-of-the-art HS based RDH methods. As a representative result, the proposed method achieves an improvement of 4.30 dB in terms of PSNR when 105,000 bits are hidden into the test Lenna image.

Digital image forgery detection by edge analysis

The advent of user-friendly yet powerful editing softwares has cast doubt on the authenticity of digital images. Therefore, developing reliable detection techniques is of great importance to verify the originality of a given image. In this work, a forgery detection technique based on the analysis of edge information is proposed. Unlike the conventional methods, the proposed technique is not restricted to the traces left by the act of double compression, but instead it allows the input image to be singly compressed or uncompressed. Experimental results confirmed that proposed method is able to localize forged area when the forged image is not double compressed.

Reversible data hiding by adaptive modification of prediction errors

Histogram Shifting (HS) is one of the most popular reversible data hiding techniques that has received tremendous attention from the research community in recent years. While histogram shifting offers many advantages, it suffers from relatively low payload, which restricts its applications significantly. In this work, a new reversible data hiding technique based on the modification of the histogram of prediction errors is proposed. The proposed method employs an adaptive strategy to vacate multiple bins as the embedding venues in order to increase the effective payload. The histogram bins are shifted dynamically based on their magnitudes. To maintain high quality for the output image, the distance of shifting is minimized for smaller prediction errors. On the other hand, the distance of shifting is allowed to be larger for larger prediction errors, which are of lower occurrences, to create more space for embedding. The proposed data hiding method is able to reversibly hide larger number of bits into the host image while achieving comparable output image quality when compared to the conventional histogram shifting based methods. The experimental results suggest that, on average, the proposed method is able to embed 0.247bpp into various standard test images, while still maintaining the visual quality at satisfactory level of ~48.9 dB.

Quality degradative reversible data embedding using pixel replacement

Conventionally, reversible data embedding methods aim at maintaining high output image quality while sacrificing carrier capacity. Recently, as a new trend, some researchers exploited reversible data embedding techniques to severely degrade image quality. In this paper, a novel high carrier capacity data embedding technique is proposed to achieve quality degradation. An efficient pixel value estimation method called checkerboard based prediction is proposed and exploited to realize data embedding while achieving scrambling effect. Here, locations of the predicted pixels are vacated to embed information while degrading the image quality. Basic performance of the proposed method is verified through experiments using various standard test images. In the best case scenario, carrier capacity of 7.31 bpp is achieved while the image is severely degraded.

A hybrid approach for multiple blastomeres identification in early human embryo images

Automatic quality assessment of the human embryo paves the way to improve the outcome of the In Vitro Fertilization (IVF) treatment by selecting embryos with the highest implantation potentials. Analyzing the shape, size, and motion of the cells, as well as other time-related changes, facilitates embryo quality assessment. However, the ambitious 3D-like side-lit appearance of the embryo, occlusion, transparency of cells and artifacts such as fragmentation make automatic detection of blastomeres (embryonic cells) a challenging task. In this paper, an automated noninvasive approach is proposed to identify multiple blastomere cells inside an embryo at different growth stages. In particular, the proposed method aims to identify up to 8 blastomeres in microscopic human embryo images of days 1-3. The proposed system is a hybrid approach that aggregates both models and features capturing global and local characteristics to locate the boundaries of each blastomere. Experimental results on a large dataset of 271 embryo images with various blastomere numbers and sizes confirm that the proposed method identifies blastomeres with average Precision, Recall, and Overall Quality of 85.9%, 85.3%, and 76.5%, respectively.

Automatic cleavage detection in H.264 sequence of human embryo development

Automated human embryo assessment/grading techniques need to be developed to enhance In Vitro Fertilization (IVF) outcome by selecting embryos with highest implantation potentials. Recently, a number of embryo assessment/grading algorithms have been proposed, however, none of these techniques are fully automatic. In addition, they generally suffer from high computational cost, since they perform extensive computational processing in the spatial domain on every single image/frame. Video compressed domain event detection could play a key role in practicalization and automation of human embryo quality assessment. In this work, a human embryo cleavage detection (CD) technique is proposed in compressed H.264 domain. By doing so, the expensive spatial domain processing is directed to where it is most needed. The performance of the proposed method is verified using two sequences of human embryo development.

An efficient sign prediction method for DCT coefficients and its application to reversible data embedding in scrambled JPEG image

In this paper, an efficient DCT sign prediction method is proposed. Unlike the conventional methods that depend on information from both spatial and frequency domains, the proposed method operates solely in the frequency domain by exploiting the pixel value patterns represented by the corresponding DCT basis vectors. In particular, each block is classified into five categories, namely, complex-pattern, complex-nonpattern, simple-smooth, simple-pattern and simple-texture, and each is treated differently using the proposed predictor. The proposed sign prediction method is then applied to realize reversible data embedding using sign information in a scrambled JPEG compressed image. This work is the first of its kind in using sign information for data embedding purposes. Basic performance of the proposed sign prediction and the proposed reversible data embedding method in scrambled JPEG image are verified using standard test images.