Investigation of the effects of image compression on the geometric quality of digital protogrammetric imagery

Master Thesis

1997

Permanent link to this Item
Authors
Supervisors
Journal Title
Link to Journal
Journal ISSN
Volume Title
Publisher
Publisher
License
Series
Abstract
We are living in a decade, where the use of digital images is becoming increasingly important. Photographs are now converted into digital form, and direct acquisition of digital images is becoming increasing important as sensors and associated electronics. Unlike images in analogue form, digital representation of images allows visual information to· be easily manipulated in useful ways. One practical problem of the digital image representation is that, it requires a very large number of bits and hence one encounters a fairly large volume of data in a digital production environment if they are stored uncompressed on the disk. With the rapid advances in sensor technology and digital electronics, the number of bits grow larger in softcopy photogrammetry, remote sensing and multimedia GIS. As a result, it is desirable to find efficient representation for digital images in order to reduce the memory required for storage, improve the data access rate from storage devices, and reduce the time required for transfer across communication channels. The component of digital image processing that deals with this problem is called image compression. Image compression is a necessity for the utilisation of large digital images in softcopy photogrammetry, remote sensing, and multimedia GIS. Numerous image Compression standards exist today with the common goal of reducing the number of bits needed to store images, and to facilitate the interchange of compressed image data between various devices and applications. JPEG image compression standard is one alternative for carrying out the image compression task. This standard was formed under the auspices ISO and CCITT for the purpose of developing an international standard for the compression and decompression of continuous-tone, still-frame, monochrome and colour images. The JPEG standard algorithm &Us into three general categories: the baseline sequential process that provides a simple and efficient algorithm for most image coding applications, the extended DCT-based process that allows the baseline system to satisfy a broader range of applications, and an independent lossless process for application demanding that type of compression. This thesis experimentally investigates the geometric degradations resulting from lossy JPEG compression on photogrammetric imagery at various levels of quality factors. The effects and the suitability of JPEG lossy image compression on industrial photogrammetric imagery are investigated. Examples are drawn from the extraction of targets in close-range photogrammetric imagery. In the experiments, the JPEG was used to compress and decompress a set of test images. The algorithm has been tested on digital images containing various levels of entropy (a measure of information content of an image) with different image capture capabilities. Residual data was obtained by taking the pixel-by-pixel difference between the original data and the reconstructed data. The image quality measure, root mean square (rms) error of the residual was used as a quality measure to judge the quality of images produced by JPEG(DCT-based) image compression technique. Two techniques, TIFF (IZW) compression and JPEG(DCT-based) compression are compared with respect to compression ratios achieved. JPEG(DCT-based) yields better compression ratios, and it seems to be a good choice for image compression. Further in the investigation, it is found out that, for grey-scale images, the best compression ratios were obtained when the quality factors between 60 and 90 were used (i.e., at a compression ratio of 1:10 to 1:20). At these quality factors the reconstructed data has virtually no degradation in the visual and geometric quality for the application at hand. Recently, many fast and efficient image file formats have also been developed to store, organise and display images in an efficient way. Almost every image file format incorporates some kind of compression method to manage data within common place networks and storage devices. The current major file formats used in softcopy photogrammetry, remote sensing and · multimedia GIS. were also investigated. It was also found out that the choice of a particular image file format for a given application generally involves several interdependent considerations including quality; flexibility; computation; storage, or transmission. The suitability of a file format for a given purpose is · best determined by knowing its original purpose. Some of these are widely used (e.g., TIFF, JPEG) and serve as exchange formats. Others are adapted to the needs of particular applications or particular operating systems.
Description

Reference:

Collections