Image Data Compression Research Articles

Adaptive Lossless Image Data Compression Method Inferring Data Entropy by Applying Deep Neural Network

When we compress a large amount of data, we face the problem of the time it takes to compress it. Moreover, we cannot predict how effective the compression performance will be. Therefore, we are not able to choose the best algorithm to compress the data to its minimum size. According to the Kolmogorov complexity, the compression performances of the algorithms implemented in the available compression programs in the system differ. Thus, it is impossible to deliberately select the best compression program before we try the compression operation. From this background, this paper proposes a method with a principal component analysis (PCA) and a deep neural network (DNN) to predict the entropy of data to be compressed. The method infers an appropriate compression program in the system for each data block of the input data and achieves a good compression ratio without trying to compress the entire amount of data at once. This paper especially focuses on lossless compression for image data, focusing on the image blocks. Through experimental evaluation, this paper shows the reasonable compression performance when the proposed method is applied rather than when a compression program randomly selected is applied to the entire dataset.

Lossless medical image compression based on anatomical information and deep neural networks

Modern imaging modalities generate large volumes of medical data that place a heavy burden on both storage and transmission. Consequently, image data compression is a key research topic in the field of medical imaging. This paper proposes a compression technique that combines anatomical information and a custom deep neural network model to facilitate the transmission and storage of medical image data. This technique first divides the medical image data into specific regions based on anatomical features, namely, image density, relative position, and organ size. A deep neural network is then trained to generate a series of optimal predictors for each region. These predictors can be adaptively switched based on the characteristics of the region being compressed. The residuals are finally compressed with an entropy coding scheme. An evaluation of this compression technique shows that the proposed “divide and conquer” method indeed achieves high prediction accuracy and obtains a compression performance that is better than that of JPEG2000 by 38%.

Image data compression technology of smart grid operation based on deep learning

Image data compression technology of smart grid operation based on deep learning

Dimensionality Reduction of High-throughput Phenotyping Data in Cotton Fields

In this paper, we implement data reduction methods to reduce the size of image datasets from cotton fields for use in a high-throughput phenotyping (HTP) pipeline in order to allow for data transfer more quickly over poor internet connections. We investigate dimensionality reduction methods to accomplish this goal. Specifically, we utilize Principal Component Analysis (PCA) to compress image data into a smaller dimension space, which when uncompressed retains significant variability from the original image. To demonstrate the ability of PCA to produce quality reconstructions, we consider the example use case of detecting cotton bloom flowering patterns with reconstructed images. We employ Open Source Computer Vision (OpenCV) to generate pixel-wise masks which both further reduces the byte size of data and successfully identifies cotton bloom flowering. The results indicate a high amount of data reduction from the original to the reconstructed images; byte sizes reduce 93% through PCA while preserving around 98% variance when using a much smaller number of components. Bitwise masking with OpenCV yields a 99% reduction in file size. The results demonstrate great potential in employing machine learning techniques for the data reduction pre-processing step prior to performing subsequent analysis. This data reduction is a crucial step in developing a field-based HTP big data pipeline.

Non-Uniform Non-Orthogonal Multicarrier Underwater Communication for Compressed Sonar Image Data Transmission

Efficient and reliable underwater sonar image transmission is increasingly desired for underwater investigation by autonomous underwater vehicles (AUV). However, it is extremely challenging due to the severe interference nature of underwater acoustic channels and the large data size of sonar images. In this paper, we first combine multicarrier modulation techniques with characteristics of compressed sonar image data to improve the transmission performance, where in previous research, image compression and multicarrier modulation were investigated separately. The compressed bitstream of sonar images is regarded as a new type of image with unique frequency characteristic. A non-uniform non-orthogonal multicarrier modulation scheme based on frequency characteristics of compressed sonar image data (F-NNMC) is proposed to meet the requirements of efficient and reliable sonar image transmission. The non-orthogonality between subcarriers improves the spectrum efficiency, and the non-uniform subcarrier structure can carry more effective information of the transmitted compressed data to improve the reliability. The simulation and experiment show that the proposed F-NNMC needs less frequency resource and has better peak signal-to-noise ratio (PSNR), structural similarity (SSIM), as well as lower bit error rate (BER) compared to orthogonal and non-orthogonal modulations.

The data conundrum: compression of automotive imaging data and deep neural network based perception

Video compression in automated vehicles and advanced driving assistance systems is of utmost importance to deal with the challenge of transmitting and processing the vast amount of video data generated per second by the sensor suite which is needed to support robust situational awareness. The objective of this paper is to demonstrate that video compression can be optimised based on the perception system that will utilise the data. We have considered the deployment of deep neural networks to implement object (i.e. vehicle) detection based on compressed video camera data extracted from the KITTI MoSeg dataset. Preliminary results indicate that re-training the neural network with M-JPEG compressed videos can improve the detection performance with compressed and uncompressed transmitted data, improving recalls and precision by up to 4% with respect to re-training with uncompressed data.

Comparison Study for Three Compression Techniques (Wavelet, Contourlet and Curvelet Transformation)

Researches and studies on compressing digital images are aiming to make it easier to deal with networks, communications and Internet by reducing the size of the multimedia files transferred, and reducing the execution time and transmission time. In this research, the lossy compression method was adopted as one of the solutions that reduce the size of the data required to compress the image, through the process of compression of digital image data using Discrete Wavelet Transform algorithms using Haar filter, and Contourlet. Using Laplace and Directional Filter, Curvelet transformation using FDCT- Wrapping Technology .The performance of the algorithms used in the proposed research is also evaluated using a Ratio Compression (RC) scale, As well as the Peak signal to noise ratio (PSNR) scale, the mean sequence error (MSE) scale, the signal to noise ratio (SNR) scale, and finally, the Normalization correlation (NC) scale. Correspondence between the original image and the recovered image after compression, in order to choose the best algorithm that achieves the best compression ratio of the image and maintains the parameters of the recovered image based on the standards (MSE, PSNR, SNR, COR and CR) used with the three algorithms, and the results showed that the Curvelet transformation algorithm achieved : best compression ratio, but at the expense of image quality.

High-precision recognition of wheat mildew degree based on colorimetric sensor technique combined with multivariate analysis

To achieve in-situ non-destructive monitoring of grain mildew degree and ensure food safety, this study took wheat as the object and carried out high-precision qualitative identification of wheat mildew degree based on colorimetric sensor technique. The gas chromatography-mass spectrometry (GC–MS) technique was used to analyze the volatile components of wheat samples with different levels of mold, and to determine the components and contents of indicative volatile substances. Accordingly, we choose 12 kinds of color materials which are sensitive to specific color reaction to prepare a set of colorimetric sensors. The odor information of wheat samples with different degrees of mildew was captured using the colorimetric sensor and display it in imaging. The principal component analysis (PCA) was performed on the color feature components of the preprocessed sensor difference image to achieve compression of sensor image data and feature reduction. Different linear (KNN; LDA) and non-linear (ELM; SVM) chemometric methods were used to create a high-quality qualitative identification models for wheat mildew based on sensor image features, and in the process of model calibration, the best parameters and the quantity of principal components (PCs) of the model are determined by the five-fold cross-validation method. Based on final results, the SVM identification model achieved a 100% correct identification rate for independent samples. The results of this study show that it is viable to monitor wheat mildew degree with high precision by using the colorimetric sensor technology with strong specificity combined with appropriate stoichiometry.

COMPUTATIONAL 2D and 3D MEDICAL IMAGE DATA COMPRESSION MODELS.

In this world of big data, the development and exploitation of medical technology is vastly increasing and especially in big biomedical imaging modalities available across medicine. At the same instant, acquisition, processing, storing and transmission of such huge medical data requires efficient and robust data compression models. Over the last two decades, numerous compression mechanisms, techniques and algorithms were proposed by many researchers. This work provides a detailed status of these existing computational compression methods for medical imaging data. Appropriate classification, performance metrics, practical issues and challenges in enhancing the two dimensional (2D) and three dimensional (3D) medical image compression arena are reviewed in detail.

A New Proposed Method For A Statistical Rules-Based Digital Image Compression

Image compression depends on data compression of digital images. Its central objective is to decrease the redundancy of the image data for reducing space and the cost of transmitting data in public communication channels. This research suggests a new compression technique based on the statistical rules. The proposed technique is one of the lossy compression techniques is based upon the statistics of the pixel values of the gray-scale image. The quality of these compressed images have been evaluated using some factors like the Image size before/after the compression process, Compression Ratio, (CR), and Peak Signal to Noise Ratio, (PSNR), Mean Square Error (MSE), and Mean Space Saving (MSS). Experimental results have demonstrated that the proposed technique provides sufficient higher compression with minimal to lose data.

Image Data Compression Based on Two Hybrids Algorithms

Image compression is a way to reduce the storage space of images. Therefore, image compression process increases the performance of such as the data transmission process. This paper aims to present a new technique to compress digital image data. In this technique, two-hybrid algorithms were used to compress the image data. The first system consists of one-dimensional discrete cosine transform (DCT), differential pulse code modulation (DPCM), and Huffman code for difference signal. The second hybrid system utilizes an expert system called Learning Automata (LA) to code the difference signal obtained from the first system. A compression ratio of about (10.8:1) was obtained from the first system. The second system provides a (20.6:1) compression ratio with non-noticeable impairment. The information loss is caused by a hybrid (DCT/DPCM) system, not by the LA system. The conclusion drawn is that using two-hybrid systems to compress the image data provides a high compression ratio. Furthermore, learning automata is preferable since it removes all the redundancy in the row data. However, in learning automata, a Huffman code is determined pixel by pixel which takes a long time.

Applying the Artificial Neural Networks with Multiwavelet Transform on Phoneme recognition

There are several advantages of Phoneme recognition. identification. It is easier to use speech for data entrance spoken communication for data ingress than other tools. It allows writing user-friendly data entrance exploiter-friendly data ingress programs. There are several difficulties in speech voice communication recognition. One of these difficulties is noise. Variability in speech is another problem. Even the speech of same speaker varies. The ability of artificial neural networks to generalize and optimize more quickly than some conventional algorithms algorithmic rule has been observed in different areas of research inquiry such as speech and pattern convention recognition, financial forecasting prognostication, image data compression and noise reduction simplification in signal processing. Neural networks take advantage of the redundancy incorporated in their distributed processing structures the proposed system depends on Artificial Neural Networks Network as decision making qualification algorithm to find the best match peer for the tested phonemes. Phoneme. The data used in this project are Arabic phonemes language phoneme stored as 8-bit mono infectious mononucleosis 8000Hz PCM WAVE Sound Auditory sensation file. The results showed that the accuracy of the proposed system is 98% recognizes the phonemes efficiently.

Research on Reversible Visible Watermarking Algorithms Based on Vectorization Compression Method

Abstract In current research on reversible visible watermarking algorithm, the original visible watermark image plays an important auxiliary role, and some algorithms also entirely depend on it to restore host image without any distortion. Therefore, in order to realize semi-blind reversible visible watermarking algorithm, the conventional reversible watermarking algorithm is used to embed compressed visible watermark image data into non-visible-watermarked region of host image. However, the amount of compressed image data obtained by conventional image compression algorithm is relatively large. Therefore, a method based on vectorization compression for the visible watermark image is proposed in this paper. Firstly, it performs edge detection on visible watermark image to obtain a discrete points set $\Gamma $ of vector contour curve. Then, the discrete points in $\Gamma $ are simplified by improved Douglas–Peucker algorithm, after that it obtains compressed vector contour data of visible watermark image. In addition, a reversible visible watermarking algorithm based on convolutional relief and image alpha fusion is proposed, which realizes reversible embedding of visible watermark image and lossless restoration of host image. The experimental results show that the proposed vectorization compression method has more advantages than traditional image compression algorithms, which greatly reduces the storage space of visible watermark image with high fidelity. Additionally, the embedded watermarking image has translucent 3D relief effect, and the fusion of host image and visible watermark image becomes more natural and harmonious.

An advanced image compression technique by using a coupled compression algorithms depend on different wavelet methods

Digital images need a large storage capacity, and as a result they need a large bandwidth for data transmission to deliver to the desired destination over the network. Image compression technologies not only reduce the size of stored data, but also maintain as much as possible the output image quality. In the proposed research we review a technique for image compression that uses a distinct two-stage image encoding method using different compression algorithms and wavelet transform methods, which combines two types of effective compression algorithms that give more ability to compress image data. The proposed compression technique which coupled two image compression algorithms that put to use premium characteristics from each algorithm. The wavelet transform methods contribute effectively to finding suitable solutions to supply better compression ratios for images with high resolution. The complete series of compression includes repeated stages of encoding and decoding, in addition to the wavelet processing itself. This study will have carried out an advanced compression technique that contain a coupled compression algorithms relying on the preferred wavelets to this work from practical experiments they are, biorthogonal and Haar wavelet transform, the performance metrics for tested true HD color image will be studied. The challenge for image compression algorithms is to detect a best solution between a low compression ratio and good visual perception results. An essential measure of achieved image compression process is taken by compression ratio CR and the ratio of bit-per-pixel BPP. The CR and BPP metrics are important components in image compression techniques. Through the results of the image compression metrics in two stages, the best practical results were obtained when the compression ratio metric CR was equal to 2.3%, and this metric indicates that the compressed image can be stored using 2.3% of the original image data size. While the BPP which represent the bit number that used to store one pixel of true color image is equal to 0.575.

An advanced image compression technique by using a coupled compression algorithms depend on different wavelet methods

Digital images need a large storage capacity, and as a result they need a large bandwidth for data transmission to deliver to the desired destination over the network. Image compression technologies not only reduce the size of stored data, but also maintain as much as possible the output image quality. In the proposed research we review a technique for image compression that uses a distinct two-stage image encoding method using different compression algorithms and wavelet transform methods, which combines two types of effective compression algorithms that give more ability to compress image data. The proposed compression technique which coupled two image compression algorithms that put to use premium characteristics from each algorithm. The wavelet transform methods contribute effectively to finding suitable solutions to supply better compression ratios for images with high resolution. The complete series of compression includes repeated stages of encoding and decoding, in addition to the wavelet processing itself. This study will have carried out an advanced compression technique that contain a coupled compression algorithms relying on the preferred wavelets to this work from practical experiments they are, biorthogonal and Haar wavelet transform, the performance metrics for tested true HD color image will be studied. The challenge for image compression algorithms is to detect a best solution between a low compression ratio and good visual perception results. An essential measure of achieved image compression process is taken by compression ratio CR and the ratio of bit-per-pixel BPP. The CR and BPP metrics are important components in image compression techniques. Through the results of the image compression metrics in two stages, the best practical results were obtained when the compression ratio metric CR was equal to 2.3%, and this metric indicates that the compressed image can be stored using 2.3% of the original image data size. While the BPP which represent the bit number that used to store one pixel of true color image is equal to 0.575.

МЕТОД СТИСНЕННЯ ДАНИХ ДИФРАКЦІЙНИХ ТА ІНТЕРФЕРЕНЦІЙНИХ ЗОБРАЖЕНЬ

In the paper the diffraction and interference images received by numerical simulation and experimentally in solving fundamental and applied problems of photonics. Images are structures with a special intensity distribution formed by the initial field and the optical system. To increase the speed of processing of the image data compression method with further implementation in databases is developed. The method of diffraction and interference images compression is based on the intensity quantization. An algorithm for image quantization has been developed: target intensity values have been determined, which allow setting quantization levels, and data visualization techniques, which determine the threshold values for these levels. The algorithm also contains image segmentation by the size of the minimum size of the topological object. The vicinity of topological object is defined under the conditions of a visual registration form and do not crossing with other regions. The topological objects of the diffraction field determine as the maximum, minimum and zero intensity, and in the interference pattern such topological objects are the maximum, minimum and the region of the band splitting. Important parameters are the average value of the intensity of the whole image - which highlights its overall structure and the average value of the intensity of the local segment. The following results were obtained by compressing data from an 8-bit image in grayscale to 2 bits of color depth are enough for an interference image, and to 3 bits, which are enough for a diffraction image. The quantization differences for diffraction images and interference patterns are shown. Data compression ratios are calculated. On the one hand, the application of the obtained results and recommendations is possible in various fields of medicine, biology and pharmacy , which use laser technology, and on the other hand in the development of separate IT identification of topological objects in the light field, optical image processing and decision support in optical problems.

Design of Image Compression Storage System and Key Algorithm for Mars Rover

Focused on the image data storage, compression and transmission of mars rover, system design and key hardware/software architecture design are carried out in this paper based on the analysis of function requirements and key design features. Image compression algorithm and transmission protocol are proposed, and the image storage and compression module in Mars rover is designed and implemented. The multi-camera partition parallel image file system based on FPGA（Field Programmable Gate Array）is realized. The adaptive first 1 run length coding algorithm for image compression based on DSP is realized on this module. Finally the function of multi-camera image data storage and management, flexible control of image compression ratio, quality progressive transmission, coding of region of interest, anti-error code diffusion and image thumbnail downward transmission are realized in the engineering development, meeting the reliable, efficient and flexible application requirements of China’s first Mars exploration mission.

A novel lossless compression encoding framework for SAR remote sensing images

The resolution of SAR (synthetic-aperture radar) remote sensing images becomes higher to provide more details, but these images contain more data, which creates a limitation in terms of transport and storage. Most of existing image data compression frameworks are lossy or designed for spectral images. In this paper, we propose a novel lossless compression encoding framework for SAR remote sensing images. In the proposed framework, an outline of the image and the high-frequency components are separated and processed separately to increase the relativity of adjacent pixels. So the accuracy of prediction is improved, which makes the data compression more effective. The outline image is down-sampled to reduce data size, and the nonlocally centralized sparse representation-based super-resolution method is used to predict pixel values using the information in nonlocal similar regions. The proposed framework is evaluated with the ground range detected and PauliRGB images captured by SAR satellites. The results show that the proposed technique can get an efficient compression performance and it outperforms existing lossless compression frameworks in terms of compression efficiency.

A combined HMM-PCNN model in the contourlet domain for image data compression.

Multiscale geometric analysis (MGA) is not only characterized by multi-resolution, time-frequency localization, multidirectionality and anisotropy, but also outdoes the limitations of wavelet transform in representing high-dimensional singular data such as edges and contours. Therefore, researchers have been exploring new MGA-based image compression standards rather than the JPEG2000 standard. However, due to the difference in terms of the data structure, redundancy and decorrelation between wavelet and MGA, as well as the complexity of the coding scheme, so far, no definitive researches have been reported on the MGA-based image coding schemes. In addressing this problem, this paper proposes an image data compression approach using the hidden Markov model (HMM)/pulse-coupled neural network (PCNN) model in the contourlet domain. First, a sparse decomposition of an image was performed using a contourlet transform to obtain the coefficients that show the multiscale and multidirectional characteristics. An HMM was then adopted to establish links between coefficients in neighboring subbands of different levels and directions. An Expectation-Maximization (EM) algorithm was also adopted in training the HMM in order to estimate the state probability matrix, which maintains the same structure of the contourlet decomposition coefficients. In addition, each state probability can be classified by the PCNN based on the state probability distribution. Experimental results show that the HMM/PCNN -contourlet model proposed in this paper leads to better compression performance and offer a more flexible encoding scheme.

A 13.3 Gbps 9/7M Discrete Wavelet Transform for CCSDS 122.0-B-1 Image Data Compression on a Space-Grade SRAM FPGA

Remote sensing is recognized as a cornerstone monitoring technology. The latest high-resolution and high-speed spaceborne imagers provide an explosive growth in data volume and instrument data rates in the range of several Gbps. This competes with the limited on-board storage resources and downlink bandwidth, making image data compression a mission-critical on-board processing task. The Consultative Committee for Space Data Systems (CCSDS) Image Data Compression (IDC) standard CCSDS-122.0-B-1 is a transform-based 2D image compression algorithm designed specifically for use on-board a space platform. In this paper, we introduce a high-performance architecture for a key-part of the CCSDS-IDC algorithm, the 9/7M Integer Discrete Wavelet Transform (DWT). The proposed parallel architecture achieves 2 samples/cycle while the very deep pipeline enables very high clock frequencies. Moreover, it exploits elastic pipeline principles to provide modularity, latency insensitivity and distributed control. The implementation of the proposed architecture on a Xilinx Kintex Ultrascale XQRKU060 space-grade SRAM FPGA achieves state-of-the-art throughput performance of 831 MSamples/s (13.3 Gbps @ 16bpp) allowing seamless integration with next-generation high-speed imagers and on-board data handling networking technology. To the best of our knowledge, this is the fastest implementation of the 9/7M Integer DWT on a space-grade FPGA, outperforming previous implementations.