Kazuki Minemura

JPEG image scrambling without expansion in bitstream size

In this work, an algorithm is proposed to scramble an JPEG compressed image without causing bitstream size expansion. The causes of bitstream size expansion in the existing scrambling methods are first identified. Three recommendations on AC coefficients in the scrambled image are proposed to combat unauthorized viewing. As the first step of the scrambling algorithm, edges are identified directly in the frequency domain using solely AC coefficients without relying on any traditional methods. These edges then form a low resolution image of its original counterpart and the information is utilized to identify regions. The DC coefficients are encoded in region-basis to suppress bitstream size expansion while achieving scrambling effect. Experiments were carried out to verify the basic performance of the proposed scrambling method. For the parameter settings considered, most of the scrambled images are of smaller bitstream size than their original counter parts.

Progressive quality degradation in JPEG compressed image using DC block orientation with rewritable data embedding functionality

This paper proposes a novel block rotational method to degrade quality and embed external data in JPEG compressed image. The orientation of each non-overlapping DC coefficients block is exploited to embed information while introducing distortion. To achieve progressive quality degradation, size of DC coefficients block is manipulated and the proposed embedding process is applied recursively by shrinking block size in each iteration. Markers are added into the blocks as pre-processing steps to ensure that the original orientation always yields the smallest difference. A post-processing is also proposed to erase the marker introduced for recovering image at higher quality, making the proposed method a rewritable method but not complete reversible. Experiments are conducted to verify the basic performance of the proposed method and comparisons with the conventional methods are also carried out.

A scrambling framework for block transform compressed image

In this work, we propose a scrambling framework for block transform compressed image. First, three attacks are proposed to sketch the outline of the original image directly from its scrambled counterpart by exploiting information deduced from the transformed components. Based on the proposed sketch attacks, a scrambling framework aiming to minimize the bitstream size overhead and prevent the leakage of visual information is put forward. In particular, the DC components are manipulated within each non-overlapping region to achieve the scrambling while simultaneously reducing the bitstream size overhead. The non-DC components are shuffled and substituted to generate a completely distorted image while preventing information leakage. The ideas are implemented in JPEG to verify its performance and compare to that of the conventional JPEG based scrambling methods. Results indicate that the proposed methods exhibit stable performance in terms of the bitstream size overhead when using different quality factors, and it is able to withstand the proposed sketch attacks as well as the classical cryptographic attacks.

HEVC selective encryption using transform skip signal and sign bin

A selective video encryption method based on the manipulation of transform skip signal and sign bin is proposed for the HEVC standard. The basic performance of the proposed selective video encryption method is evaluated in terms of perceptual inspection, outline detection and sketch attack using various classes of test video sequences. Preliminary results show that the proposed method provides quality degradation up to -0.22 in SSIM score when compared to the conventional method [5]. In addition, the edge difference ratio is greater than 0.73, which is closed to the perfect dissimilarity with respect to the original video. Functional comparison between the proposed method and the conventional selective encryption methods is then presented.

Multi-oriented text detection for intra-frame in H.264/AVC video

Text detection in compressed video has received much attention in recent years due to the effectiveness of DCT coefficients and motion vectors in realizing several applications. In this paper, a new text detection, which utilizes AC coefficients in the H.264/AVC compressed video, is proposed. The proposed median deviation of coefficients from a specific subband is first computed, then the k-means clustering and morphological operations are applied to classify the text candidates. The majority orientation is considered to eliminate false positive candidate groups that have different orientations. Local block energy information is extracted to obtain the final text candidates. Experimental results show that the proposed method outperforms the existing methods either in computational time or accuracy in detecting horizontal text. Furthermore, for non-horizontal text, the proposed method is superior to all the conventional methods considered.

Sketch attacks: A note on designing video encryption method in H.264/AVC

In this work, we propose four sketch attacks on H.264/AVC encrypted-compressed video. First, we briefly describe the notion of sketch attack, then deploy the conventional sketch attacks, which are designed for still image, to sketch the frames of H.264/AVC compressed video. Next, we propose four sketch attacks to generate outline of the original frame by using partially decoded information of the H.264/AVC compressed video, including residue DC coefficients, residue AC coefficients, motion vectors, and macroblock bitstream size. Experiments are conducted to verify the performance of the proposed sketch attacks using the ICADR2013 and HEVC dash video data sets. The proposed sketch attacks are also compared with the four conventional sketch attacks. Results indicate that the proposed sketch attack can generate the outline of a frame regardless of its type, where the fidelity of the sketch image improves as resolution increases.

A Novel Sketch Attack for H.264/AVC Format-Compliant Encrypted Video

In this paper, we propose a novel sketch attack for H.264 advanced video coding (H.264/AVC) format-compliant encrypted video. We briefly describe the notion of sketch attack, review the conventional sketch attacks designed for discrete cosine transform (DCT)-based compressed image, and identify their shortcomings when applied to attack compressed video. Specifically, the conventional DCT-based sketch attacks are incapable in sketching outlines for inter frame, which is deployed to significantly reduce temporal redundancy in video compression. To sketch directly from inter frame, we put forward a sketch attack by considering the partially decoded information of the H.264/AVC compressed video, namely, the number of bits spent on coding a macroblock. To evaluate the sketch image, we consider the Canny edge map as the ideal outline image. Experiments are conducted to verify the performance of the proposed sketch attack using ICADR2013, High Efficiency Video Coding dash, and Xiph video data sets. Results suggest that the proposed sketch attack can generate the outline image of the original frame for not only intra frame but also inter frame.

Moving object detection in HEVC video by frame sub-sampling

Video compression aims to remove spatial-temporal redundancies where the encoded bitstream, particularly the motion vectors, may not represent the actual motions in the video. Hence, moving object detection in the compressed video stream is a technically challenging task. In this work, we propose a novel moving object detection algorithm using frame sub-sampling method in the state-of-the-art HEVC video coding standard. Specifically, the number of frames is reduced by means of (temporal) sub-sampling. The frames are re-encoded using HEVC with the same environmental setting to amplify the motion of the moving objects. Sub-sampling effectively increases the motion intensity of the objects, which can be the significant cue for detecting moving object while motions in the background still remain small. Motion vectors and INTRA coding units of moving object obtained via frame sub-sampling and re-encoding are selectively utilized to separate the background and moving objects in the video. The segmented results are refined and compared with the result without performing frame sub-sampling. Results show that the sub-sampling method achieves higher accuracy, with an improvement greater than 0.35 in terms of F-measure.

A Rewritable Data Embedding in JPEG-XR Compressed Image with File Size Preservation

In this paper, we propose a rewritable data embedding method in the JPEG-XR compressed domain. Our proposed method is based on modifying the sign of the selected FCT (Forward Core Transform) coef- ficients. Specifically, the last coefficient, with respect to the conventional zigzag scanning order, in a 4 × 4 HP block is selected to embed data. Af- ter extracting the embedded data, the proposed method can restore the image quality by using the proposed simple majority vote method. Al- though the restoration does not perfectly regenerate the original image, the proposed data embedding method is rewritable. To evaluate the pro- posed method, four criteria including quality of the output image (i.e., embedded with data), file size of the output image, available payload and reconstructed image quality, are considered. Experiments are carried out to verify the basic performance of the proposed method. In the best case scenario, the proposed method can embed up to 15954 bits into an 8-bit grayscale image of dimension 512× 512 pixels with a reconstructed image of SSIM = 0.9919.