Iterative Weighted Recovery for Block-Based Compressive Sensing of Image/Video at a Low Subrate

Dinh, Khanh Quoc,Jeon, Byeungwoo IEEE 2017 IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDE Vol.27 No.11

<P>In compressive sensing (CS) of images or videos, a block-based sensing or recovery scheme can facilitate low-cost sampling or recovery in memory and computation. However, its recovery with small block size and small subrate suffers greatly from its lack of information of the measurement data essential to recover a unique solution among many candidates. This study, based on prior knowledge of the signal to be sensed, namely, the relative magnitude difference of signal entries, designs a weighting process to limit the solution space of the recovered signal and combines it with much simplified Landweber iterations to deliver a complete recovery algorithm, called iterative weighted recovery (IWR). We theoretically verify the performance of the proposed IWR, including error bound, convergence rate, and stopping criterion. Application of the proposed IWR to block-based CS of images or videos confirms the quality improvement of the recovered images or videos and reduction of recovery time.</P>

Weighted overlapped recovery for blocking artefacts reduction in block-based compressive sensing of images

Khanh Quoc Dinh,Hiuk Jae Shim,Byeungwoo Jeon IET 2015 Electronics letters Vol.51 No.1

<P>In compressive sensing (CS) of images, a block-based framework is preferred to avoid the huge memory and computation required for a frame-based approach. However, the recovered image suffers from blocking artefacts due to independent block processing, especially at a low subrate. As a result of this reported work the artifacts are reduced by weighted averaging adopting two techniques: overlapped CS recovery and adaptive weighting. Simulation results show its improvement in both subjective and objective qualities.</P>

Measurement Coding for Compressive Sensing of Color Images

Dinh, Khanh Quoc,Trinh, Chien Van,Nguyen, Viet Anh,Park, Younghyeon,Jeon, Byeungwoo The Institute of Electronics and Information Engin 2014 IEIE Transactions on Smart Processing & Computing Vol.3 No.1

From the perspective of reducing the sampling cost of color images at high resolution, block-based compressive sensing (CS) has attracted considerable attention as a promising alternative to conventional Nyquist/Shannon sampling. On the other hand, for storing/transmitting applications, CS requires a very efficient way of representing the measurement data in terms of data volume. This paper addresses this problem by developing a measurement-coding method with the proposed customized Huffman coding. In addition, by noting the difference in visual importance between the luma and chroma channels, this paper proposes measurement coding in YCbCr space rather than in conventional RGB color space for better rate allocation. Furthermore, as the proper use of the image property in pursuing smoothness improves the CS recovery, this paper proposes the integration of a low pass filter to the CS recovery of color images, which is the block-based ${\ell}_{20}$-norm minimization. The proposed coding scheme shows considerable gain compared to conventional measurement coding.

Reliability-Based Deblocking Filter for Wyner-Ziv Video Coding

Khanh Quoc Dinh,Hiuk Jae Shim,Byeungwoo Jeon 대한전자공학회 2016 IEIE Transactions on Smart Processing & Computing Vol.5 No.2

In Wyner-Ziv coding, video signals are reconstructed by correcting side information generated by block-based motion estimation/compensation at the decoder. The correction is not always accurate due to the limited number of parity bits and early stopping of low-density parity check accumulate (LDPCA) decoding in distributed video coding, or due to the limited number of measurements in distributed compressive video sensing. The blocking artifacts caused by blockbased processing are usually conspicuous in smooth areas and degrade the perceptual quality of the reconstructed video. Conventional deblocking filters try to remove the artifacts by treating both sides of the block boundary equally; however, coding errors generated by block-based processing are not necessarily the same on both sides of the block boundaries. Such a block-wise difference is exploited in this paper to improve deblocking for Wyner-Ziv frameworks by designing a filter where the deblocking strength at each block can be non-identical, depending on the reliability of the reconstructed pixels. Test results show that the proposed filter not only improves subjective quality by reducing the coding artifacts considerably, but also gains rate distortion performance.

Color Image Denoising via Cross-Channel Texture Transferring

Dinh, Khanh Quoc,Canh, Thuong Nguyen,Jeon, Byeungwoo IEEE 2016 IEEE signal processing letters Vol.23 No.8

<P>Image denoising can reduce the perturbation inevitably generated during image signal acquisition and its subsequent processing. While the utilization of nonlocal properties can enhance the performance of the state-of-the-art denoising methods, a heavy computational burden is incurred especially for color images. Inspired by the high correlation in the texture information over color channels, for a reduction of the computational burden, this letter proposes denoising the luma channel first, and then, performing a patch-wise linear prediction to transfer the texture information of the denoised luma channel to the other two channels. The texture transferring is adapted to local characteristic (i.e., variance of the local patches) for a reduction of color smearing caused by large prediction error especially along edges. Experimental results confirm that the proposed method achieves performance improvement over the state-of-the-art color image denoising methods only at a slightly increased complexity of single-channel denoising.</P>

Small-block sensing and larger-block recovery in block-based compressive sensing of images

Dinh, Khanh Quoc,Shim, Hiuk Jae,Jeon, Byeungwoo Elsevier 2017 SIGNAL PROCESSING-IMAGE COMMUNICATION - Vol.55 No.-

<P><B>Abstract</B></P> <P>In the block-based compressive sensing (CS) of images, a small block is more practical due to its low-cost sensing in terms of the required memory and the computational complexity. A large block, however, is more effective in CS recovery because of the high probability of a smaller mutual coherence and a more-compressible representation of the images. This paper proposes a block-based CS scheme that is applicable to images with a small-block sensing and larger-block recovery (SBS-LBR), whereby a block-diagonal sensing matrix is used to arbitrarily set a recovery-block size that is multiple-times larger than the sensing block size; subsequently, a more compressible transform signal is generated with large-sized sparsifying basis. The proposed SBS-LBR not only facilitates a low sampling cost, but also improves the recovered images from the larger recovery-block size. Our experiment results confirm a theoretical analysis of the scheme, and have shown the improvement from the proposed SBS-LBR with the suggested proper choices regarding the sensing- and recovery-block sizes.</P> <P><B>Highlights</B></P> <P> <UL> <LI> ACS scheme employs different block sizes in the sensing and the recovery. </LI> <LI> The scheme is analyzed with mutual coherence and compressibility of transform signal. </LI> <LI> The scheme achieves both low sampling cost and more-favorable recovery performance. </LI> </UL> </P>

Smoothed Group-Sparsity Iterative Hard Thresholding Recovery for Compressive Sensing of Color Image

Viet Anh Nguyen,Khanh Quoc Dinh,Chien Van Trinh,Younghyeon Park(박영현),Byeungwoo Jeon(전병우) 대한전자공학회 2014 전자공학회논문지 Vol.51 No.4

압축센싱은 성긴(Sparse) 또는 압축가능한(Compressible) 신호에 대해 Nyquist rate 미만의 샘플링으로도 신호 복원이 가능하다는 것을 수학적으로 증명한 새로운 패러다임의 신호 획득 방법이다. 단순한 신호 획득 과정을 이용하면서도, 동시에 우수한 압축센싱 복원 영상을 얻기 위한 많은 연구들이 수행되고 있다. 그러나, 에너지 분포 및 인간 시각 시스템 등 컬러 영상에 대한 기본적인 특성을 복원 과정에 활용한 기존 압축센싱 관련 연구는 많이 부족하다. 이러한 문제를 해결하기 위해, 본 논문에서는 컬러영상의 압축센싱 복원을 위한 평활 그룹-희소성 기반 반복적 경성 임계 알고리즘을 제안한다. 제안하는 방법은 그룹-희소성에 기반한 경성 임계치 적용과 프레임 기반 필터의 사용을 통해 영상의 변환 영역에 대한 희소성을 증대시키는 동시에 화소 영역의 평활 정도를 복원 과정에 활용할 수 있도록 한다. 또한, 그룹-희소화 경성 임계 과정은 자연 영상의 에너지 분포 및 인간 시각 시스템 특성에 따라 중요하다고 판단되는 RGB-그룹 계수들을 보전하도록 설계하였다. 실험 결과 객관적 화질 측면에서 제안방법이 대표적인 그룹-희소화 평활 복원 기법 보다 평균 PSNR이 최대 2.7dB 높은 것을 확인하였다. Compressive sensing is a new signal acquisition paradigm that enables sparse/compressible signal to be sampled under the Nyquist-rate. To fully benefit from its much simplified acquisition process, huge efforts have been made on improving the performance of compressive sensing recovery. However, concerning color images, compressive sensing recovery lacks in addressing image characteristics like energy distribution or human visual system. In order to overcome the problem, this paper proposes a new group-sparsity hard thresholding process by preserving some RGB-grouped coefficients important in both terms of energy and perceptual sensitivity. Moreover, a smoothed group-sparsity iterative hard thresholding algorithm for compressive sensing of color images is proposed by incorporating a frame-based filter with group-sparsity hard thresholding process. In this way, our proposed method not only pursues sparsity of image in transform domain but also pursues smoothness of image in spatial domain. Experimental results show average PSNR gains up to 2.7dB over the state-of-the-art group-sparsity smoothed recovery method.

Study on hybrid sensing matrix for compressive sensing of images

Minh Van Phan,Khanh Quoc Dinh,Byeungwoo Jeon(전병우) 한국방송·미디어공학회 2014 한국방송공학회 학술발표대회 논문집 Vol.2014 No.6

Compressive Sensing Recovery of Natural Images Using Smooth Residual Error Regularization

Chien Van Trinh,Khanh Quoc Dinh,Viet Anh Nguyen,Younghyeon Park(박영현),Byeungwoo Jeon(전병우) 대한전자공학회 2014 전자공학회논문지 Vol.51 No.6

압축센싱은 성긴 (sparse) 신호에 대해 Nyquist rate 미만의 샘플링으로도 신호 획득이 가능하다는 것을 수학적으로 증명한 새로운 개념이다. 그동안 영상분야 압축센싱을 위한 수많은 복원 알고리즘들이 제안되어 왔으나, 낮은 측정률 하에서는 복원 화질 측면에서 아직 개선할 점이 많다. 일례로, 자연 영상의 압축센싱 복원 화질 향상을 위해, 영상과 관련한 사전 정보들로부터 정규화 식을 도출하여 복원에 적용해 볼 수 있을 것이다. 따라서, 본 논문에서는 Dantzig selector 및 평활 필터(가우시안필터 및 nonlocal 평균 필터)기반의 평활 잔차 오류 정규화 방법을 제안한다. 또한, 복원 영상의 객체 및 배경에서 발생하는 edge 정보를 우수하게 보전하는 것으로 알려진 Total variation 기반 최소화 알고리즘에 적용하여 복원 영상의 화질을 향상시키는 방법을 제안한다. 제안하는 구조는 잔차신호의 평활화를 활용한다는 측면에서 새로운 압축센싱 복원 방식이라고 할 수 있다. 실험 결과, 제안방법은 기존 방법들에 비해 객관적 및 주관적 화질 측면에서 더 높은 성능 향상을 보여주었으며, 특히 기존 Bayesian 압축센싱 복원 방식과 비교 시 최대 9.14 dB 성능이 향상되었다. Compressive Sensing (CS) is a new signal acquisition paradigm which enables sampling under Nyquist rate for a special kind of signal called sparse signal. There are plenty of CS recovery methods but their performance are still challenging, especially at a low sub-rate. For CS recovery of natural images, regularizations exploiting some prior information can be used in order to enhance CS performance. In this context, this paper addresses improving quality of reconstructed natural images based on Dantzig selector and smooth filters (i.e., Gaussian filter and nonlocal means filter) to generate a new regularization called smooth residual error regularization. Moreover, total variation has been proved for its success in preserving edge objects and boundary of reconstructed images. Therefore, effectiveness of the proposed regularization is verified by experimenting it using augmented Lagrangian total variation minimization. This framework is considered as a new CS recovery seeking smoothness in residual images. Experimental results demonstrate significant improvement of the proposed framework over some other CS recoveries both in subjective and objective qualities. In the best case, our algorithm gains up to 9.14 dB compared with the CS recovery using Bayesian framework.

Multi-Resolution Kronecker Compressive Sensing

Canh, Thuong Nguyen,Quoc, Khanh Dinh,Jeon, Byeungwoo The Institute of Electronics and Information Engin 2014 IEIE Transactions on Smart Processing & Computing Vol.3 No.1

Compressive sensing is an emerging sampling technique which enables sampling a signal at a much lower rate than the Nyquist rate. In this paper, we propose a novel framework based on Kronecker compressive sensing that provides multi-resolution image reconstruction capability. By exploiting the relationship of the sensing matrices between low and high resolution images, the proposed method can reconstruct both high and low resolution images from a single measurement vector. Furthermore, post-processing using BM3D improves its recovery performance. The experimental results showed that the proposed scheme provides significant gains over the conventional framework with respect to the objective and subjective qualities.