JPEG Compression Technique Research Articles

OODNet: A deep blind JPEG image compression deblocking network using out-of-distribution detection

JPEG is one of the most popular image compression techniques, with numerous applications ranging from medical imaging to surveillance systems. Since JPEG introduces the blocking artifacts to the decompressed visual signals, enhancing the quality of these images is of paramount importance. Recently, various deep neural networks have been proposed for JPEG image deblocking that can effectively reduce the blocking artifacts produced by the JPEG compression technique. However, most of these schemes could only handle decompressed images generated by a set of specific JPEG quality factor (QF) values employed in the network training process. Therefore, when the images are obtained by the JPEG QF values other than those used in the network training process, the performance of deep learning-based JPEG image deblocking schemes drops significantly. To address this, in this paper, we propose a novel deep learning-based blind JPEG image deblocking method, which employs out-of-distribution detection to perform deblocking efficiently for various quality factor (QF) values. The proposed scheme can distinguish between the decompressed images using the QF values used in the training set and those using the QF values not used in the training set, and then, a suitable deblocking strategy for generating high-quality images is developed. The proposed scheme is shown to outperform the state-of-the-art JPEG image deblocking methods for various QF values.

Distributed JPEG Compression and Decompression for Big Image Data Using Map-Reduce Paradigm

Digital data is primarily created and delivered in the form of images and videos in today’s world. Storing and transmitting such a large number of images necessitates a lot of computer resources, such as storage and bandwidth. So, rather than keeping the image data as is, the data could be compressed and then stored, which saves a lot of space. Image compression is the process of removing as much redundant data from an image as feasible while retaining only the non-redundant data. In this paper, the traditional JPEG compression technique is executed in the distributed environment with map-reduce paradigm on big image data. This technique is carried out in serial as well as in parallel fashion with different number of workers in order to show the time comparisons between these setups with the self-created large image dataset. In this, more than one Lakh (121,856) images are compressed and decompressed and the execution times are compared with three different setups: single system, Map-Reduce (MR) with 2 workers and MR with 4 workers. Compression on more than one Million (1,096,704) images using single system and MR with 4 workers is also done. To evaluate the efficiency of JPEG technique, two performance measures such as Compression Ratio (CR) and Peak Signal to Noise Ratio (PSNR) are used.

Usage analysis of SVD, DWT and JPEG compression methods for image compression

Through image compression, can save bandwidth usage on telecommunication networks, accelerate image file sending time and can save memory in image file storage. Technique to reduce image size through compression techniques is needed. Image compression is one of the image processing techniques performed on digital images with the aim of reducing the redundancy of the data contained in the image so that it can be stored or transmitted efficiently. This research analyzed the results of image compression and measure the error level of the image compression results. The analysis to be carried out is in the form of an analysis of JPEG compression techniques with various types of images. The method of measuring the compression results uses the MSE and PSNR methods. Meanwhile, to determine the percentage level of compression using the compression ratio calculation. The average ratio for JPEG compression was 0.08605, the compression rate was 91.39%. The average compression ratio for the DWT method was 0.133090833, the compression rate was 86.69%. The average compression ratio of the SVD method was 0.101938833 and the compression rate was 89.80%.

Image forgery detection based on statistical features of block DCT coefficients

Abstract Majority of the existing detection algorithms are able to deal with either type of forgery (splicing or copy-move). However, if we require a unified approach, we need to combine two previous works for each one of the alterations. In order to solve this problem, the authors present a new algorithm for the detection of splicing and copy-move forgery in the same instance. In this paper, a forgery detection technique is proposed which exploits the artifacts originated due to manipulations performed on JPEG encoded images. In JPEG compression technique, an image is divided into non-overlapping blocks of size 8x8 pixels and discrete cosine transform (DCT) coefficients are evaluated for each block independently. When a JPEG compressed image is tampered, there is a change in the statistical properties of AC components of block DCT coefficients. To capture this change, we propose to use standard deviation and count of non-zero DCT coefficients corresponding to each of the AC frequency components independently. The images are cropped by removing a few rows and columns from the top left corner and suggested features are evaluated for test image and its cropped version. The extracted feature vector is used with the support vector machine (SVM) for the classification of authentic and forged images. Experiments are conducted on a standard dataset of pre- and post-processed forged images CASIA vl.O and v2.0 to consolidate the theoretical concept of the proposed technique. Also, the comparative analysis is performed to showcase better detection rates compared with the state-of-the-art methods.

Digital Watermarking using Discrete Wavelet Techniques with the help of Multilevel Decomposition Technique

Digital watermarking plays an important role for securing data, providing authentication, confidentiality and even protects digital images from illegal distortion. In this paper, we are particular using discrete wavelet transform which is more prevalent than others. The watermark logo is inserted after compressing the image with the help of JPEG compression technique. First original image is taken; it is changed into multilevel decomposed with the help of HAAR wavelet, as it is a type of wavelet used for diet (discrete wavelet technique) technique. The algorithm has been tested under the presence of attacks like Gaussian noise, JPEG compression and we see how their BER and PSNR value varies. Further, compression before watermarking is done to ensure inaccessibility of information to unauthorized personnel.

A Novel Medical Image Compression Technique based on Structure Reference Selection using Integer Wavelet Transform Function and PSO Algorithm

The utility of medical image increases due to serious disease prediction such as brain stroke and lung cancer. The size of medical image required large amount of memory and bandwidth for storage and transmission. For the lossless compression of medical image some standard algorithm are used such as JPEG, composite and Wavelet based. These entire algorithms perform good compression ratio and PSNR value. But the redundant structure of medical image reduces the compression ratio and decreases the value of PSNR and suffered quality of medical image compression. In this paper we proposed a structure reference selection process for collecting redundant frame of structure for compression. The collection of redundant frame structure is performed by particle swarm optimization algorithm. Particle swarm optimization is searching technique. For the generation of structure of frame we used integer wavelet transform function. The proposed algorithm is implemented through MATLAB software and is compared with composite algorithm, SOM algorithm and JPEG compression technique. General Terms Integer wavelet transforms, Reference Frame, Image Compression et. al.

Efficient DCT-based image compression technique

Although JPEG technique is considered as the most popular image compression standard, it behaves high visual degradation at low bit rates. In this paper, an efficient DCT–based image compression technique is proposed to achieve high Compression Ratio (CR) with high quality at both high and low bit rates. This technique uses switching between JPEG compression technique at high bit rates and a novel Adaptive Lossy Image Compression (ALIC) technique at low bit rates. ALIC is proposed to overcome the drawbacks of JPEG technique at low bitrates. The performance of the proposed technique is analysed at low and high bit rates on both grey and colour images. Performances of both JPEG and ALIC techniques are analysed and compared. The experimental results reveal that the proposed ALIC technique achieves better CR with acceptable SNR in comparison with JPEG technique. Also, the resultant CR of ALIC technique can be considerably increased with a slight decrease of its PSNR. This decrease in PSNR does not result in a noticeable visual degradation of the compressed image. On the other hand, increasing the CR of JPEG technique results in a noticeable visual degradation due to the appearance of blocking effect in the reconstructed image. Thus, it is greatly recommended to use ALIC technique in the applications that require high CR with stable PSNR. ALIC is a general purpose technique that can be applied, not only on images, but also on any data source which uses Huffman coding to achieve better CR. Therefore, it is suitable for compression of text, image and video.

Ultrasound Beamforming Using Compressed Data

The rapid advancements in electronics technologies have made software-based beamformers for ultrasound array imaging feasible, thus facilitating the rapid development of high-performance and potentially low-cost systems. However, one challenge to realizing a fully software-based system is transferring data from the analog front end to the software back end at rates of up to a few gigabits per second. This study investigated the use of data compression to reduce the data transfer requirements and optimize the associated trade-off with beamforming quality. JPEG and JPEG2000 compression techniques were adopted. The acoustic data of a line phantom were acquired with a 128-channel array transducer at a center frequency of 3.5 MHz, and the acoustic data of a cyst phantom were acquired with a 64-channel array transducer at a center frequency of 3.33 MHz. The receive-channel data associated with each transmit event are separated into 8 × 8 blocks and several tiles before JPEG and JPEG2000 data compression is applied, respectively. In one scheme, the compression was applied to raw RF data, while in another only the amplitude of baseband data was compressed. The maximum compression ratio of RF data compression to produce an average error of lower than 5 dB was 15 with JPEG compression and 20 with JPEG2000 compression. The image quality is higher with baseband amplitude data compression than with RF data compression; although the maximum overall compression ratio (compared with the original RF data size), which was limited by the data size of uncompressed phase data, was lower than 12, the average error in this case was lower than 1 dB when the compression ratio was lower than 8.

Image Compression using Approximate Matching and Run Length

Image compression is currently a prominent topic for both military and commercial researchers. Due to rapid growth of digital media and the subsequent need for reduced storage and to transmit the image in an effective manner Image compression is needed. Image compression attempts to reduce the number of bits required to digitally represent an image while maintaining its perceived visual quality. This study concentrates on the lossless compression of image using approximate matching technique and run length encoding. The performance of this method is compared with the available jpeg compression technique over a wide number of images, showing good agreements.

Digital watermarking method using compression concept and coefficient difference

In this paper, a robust watermarking technique for embedding a watermark into grey-level image is proposed. It is well known that lossy compression will remove high-frequency components of an image. Using the concept of compression to embed a watermark into the coefficient difference between the middle frequency of the original and the compressed images, the proposed scheme provides a robust characteristic and maintains good visual quality of an image. The watermark is designed to be extracted without the original image so that the application is more feasible in practice. Furthermore, security is strongly enhanced by the chaos technique which is used in the proposed scheme. Experimental results show that the scheme features not only imperceptibility and security, but also robustness against various image processing attacks such as JPEG and JPEG2000 lossy compression techniques, filtering, noise adding, scaling, cropping, etc.

A new approach for optical colored image compression using the JPEG standards

Image compression consists in reducing information volume representing an image. Elimination of redundancies and non-pertinent information enables memory space minimization and thus faster data transmission. The present work aims to improve the quality of the compressed image while minimizing the time required for compression by using the principle of coherent optics. We present an optical adaptation of the method of JPEG compression technique for binary, gray-level and color images. Illustrative simulations will be given at the end to validate our architecture and to evaluate the performance on different types of images (binary, gray and color). An optical implementation setup is proposed and validated experimentally.

Effects of jpeg 2000 compression on automated dsm extraction: evidence from aerial photographs

AbstractThis paper evaluates the effects of JPEG 2000 compression on automated digital surface model (DSM) extraction, using the area‐based matching technique. Two stereopairs of aerial photographs, which have different photo scale, pixel size, and topographic type, are used as the images for the experiments. In the experiments, the DSM generated from the uncompressed image is set as the reference, and the elevation accuracy is computed for a range of compression ratios (2:1 to 100:1). The results show that the employed indices of image quality based on the JPEG 2000 compression are clearly superior to the commonly used JPEG compression technique—especially for high compression ratios. However, the elevation accuracy varies with the compression ratios. Also, JPEG 2000 compression does not have a significant influence on the percentage of the successful matching rate. In addition, a near linear fall‐off is observed in extracted DSM accuracy with increasing compression ratios.

Detection and restoration of tampered JPEG compressed images

To protect the integrity of digital images, our scheme proposes a technique based on standard JPEG still image compression. The proposed scheme can detect tampered images and can also recover them. Images are compressed using the JPEG compression technique, and the proposed scheme also incorporates an edge detection technique. First, the edge detection technique identifies the edges of the image before image compression takes place. The obtained edge characteristic is then embedded into the image immediately after compression. If the image is tampered during transmission, the embedded edge characteristic can be used to detect the tampered areas, and these tampered areas can be reconstructed using the interpolation method and the embedded edges. Therefore, the proposed scheme allows recipients to know that the image has been tampered with during transmission.

Lossless compression of multispectral satellite images

This paper describes a neural network-based technique to compress multispectral SPOT satellite images losslessly. The technique harnesses the pattern recognition property of one-hidden-layer back propagation neural networks to exploit both the spatial and the spectral redundancy of the three-band SPOT images. The networks are initially trained on samples of the SPOT images with a unique network for each of the bands. The resultant trained nonlinear predictors are then used to predict the target SPOT images. Predicted errors are entropy-coded using multi-symbol arithmetic coding. This technique achieves compression ratios of 2.1 times and 3.2 times for urban and rural SPOT images respectively which are above 10% better than using lossless JPEG compression techniques. In comparison with JPEG2000 lossless compression, the proposed technique is 5% better.

A Novel Approach to Lossy Real-Time Image Compression: Hierarchical Data Reorganization on a Low-Cost Massively Parallel System

This paper discusses an innovative real-time oriented image compression system, based on a simple algorithm designed explicitly to be implemented on a low-cost SIMD computer architecture featuring a much lower power consumption than traditional DSPs or dedicated hardware. For this reason the considered approach is suitable to be integrated on portable systems, where power consumption is a critical design issue. The algorithm, based on a hierarchical decomposition of the input image, has been tested on a special purpose SIMD system, PAPRICA, exploiting its features such as its massive parallelism and its capability to operate on local data as well as to handle hierarchical data structures. According to the proposed approach, the quality of the decompressed image can be traded for a lower power consumption as well as a higher processing speed. A comparison between the discussed algorithm and the standard JPEG compression technique is also presented.

Image compression by the wavelet decomposition

AbstractDecomposition of images with the Haar orthonormal basis which is an important member of compactly supported Wavelets and a quadtree structured hierarchical coding technique are used in this work to obtain high image compression efficiency and time complexity linear in the number of pixels. An exhaustive testing of the algorithm has been performed on images of different complexity and typical of some application environment (image transmission and storing, remote control of intelligent robots). The results of the experiments arc presented and discussed. Finally, a comparison of quality performance of these techniques with the JPEG (Block Cosine Transform coding) compression technique is presented.