In this paper, a new compression technique aiming at reducing the size of storage of multispectral images and maintaining at the same time the high-quality reconstruction is presented. An optimal multispectral band ordering process is applied before compression, and then, the dual-tree discrete wavelet transform is used in the spectral dimension, and the 2D discrete wavelet transform is used in the spatial dimensions. Finally, a simple Huffman coder is used for compression. Landsat ETM+ images are used for experimentations. Experimental results demonstrate that the proposed technique has better performance than JPEG, JPEG2000, SPIHT, and JPEG2000 with a 3D dual-tree transformation.

Multispectral image compression with band ordering and wavelet transforms

The development of new technologies and communication networks creates new needs and stimulates the introduction of new functionalities particularly in high speed internet data paths. The drawbacks of traditional Discrete Wavelet Transform (DWT) in image compression were suitably estimated. This paper investigates the importance of new technique to compress X-Ray medical images and satellite extracted images. It also verified for other types of images also. The proposed method out performs the existing methods and the simulation results of the new proposed algorithm shows better performance. The values of PSNR and Compression ratios clearly show the robustness of the algorithm in compression of the images. The simulation results performed on MATLAB.

A robust approach for qualitative compression in JPEG 2000 standard

Along with the rapid development of deep learning and artificial intelligence technology, machine vision instead of human vision is used in many applications, such as defect detection, face recognition, etc. In the power system, it is very important to use computer vision algorithms to perform intelligent defect detection on images taken by Unmanned Aerial Vehicles, robots and other equipment. Traditional image compression algorithms aim to reduce the pixel distortion based on the visual characteristics of the human eye, which are not suitable for machine vision. Considering the limited communication channel bandwidth, the images generally compressed at low bitrate. Then, the blocking artifact introduced by compression will result in undetectable or false detection of the state of small electrical components, such as insulator strings, anti-vibration hammers, etc. In this paper, a quantization algorithm which is oriented to machine vision is proposed to address these problems. First of all, low-level image features are extracted based on CNN network. Subsequently, according to the extracted features, the original images are divided into key areas and non-key areas. Finally, key areas are encoded with smaller quantization level and non-key areas are encoded with lager quantization level. Experimental results demonstrate that compared with traditional image compression algorithm, the proposed algorithm improves the accuracy of defect detection from 20% to more than 80% at bitrate of 0.24bpp.

An Improved Quantization Algorithm For Electric Power Inspection

In the present era of Internet multimedia data especially Images and Videos are the most widely used digital format for data transmission. However due to their large data sizes and constraint of low bandwidth capacity of communication channel it is very difficult to transmit them at optimum speed maintaining the signal quality. Compression therefore, is a vital tool that not only reduces the data size thereby leading to faster data transmission but also protects it to some extent from transmission errors. A large variety of Image and Videos compression techniques are employed each having their own strengths and weaknesses. This paper is an effort to present an overview of image and video compression techniques, their working and comparison.

/pdf/a-comparative-study-on-image-and-video-compression-46749rfsrc.pdf

A Comparative study on Image and Video Compression Techniques

A new algorithm for the embedding of a prediction mechanism into the JPEG2000 coding chain

The JPEG algorithm for compression of grayscale and color images was implemented in software and tested on nine IS0 color test images. Image-optimized Huffman codes were generated for each image and the performance of these codes were compared with the proposed default Huffman code as well as with a code optimized for the average of the nine images. Statistics were generated for the run lengths and bits of significance required to encode the quantized DC and AC components of the Discrete Cosine Transforms (DCTs) for the Luminance and Chrominance image arrays. Statistics were also formed to determine the fraction of the code used to locate the positions of the non-zero spectral samples as opposed to the values of those samples.

https://ieeexplore.ieee.org/iel2/3982/11472/00523462.pdf

F. Harris

Papers

The JPEG Algorithm for Image Compression: A Software Implementation and some Test Results