Analysis and Comparison of Objective Methods for Image Quality Assessment

• Main Authors: P. S. Babkin, Y. N. Pavlov
• Format: Article
• Language: Russian
• Published: MGTU im. N.È. Baumana 2014-01-01
• Series: Nauka i Obrazovanie
• Subjects: algorithm; Objective methods image quality assessment; PQS; MS-SSIM
• Online Access: http://technomag.edu.ru/jour/article/view/688

<p>The purpose of this work is research and modification of the reference objective methods for image quality assessment. The ultimate goal is to obtain a modification of formal assessments that more closely corresponds to the subjective expert estimates (MOS).</p><p>In considering the formal reference objective methods for image quality assessment we used the results of other authors, which offer results and comparative analyzes of the most effective algorithms. Based on these investigations we have chosen two of the most successful algorithm for which was made a further analysis in the MATLAB 7.8 R 2009 a (PQS and MSSSIM). The publication focuses on the features of the algorithms, which have great importance in practical implementation, but are insufficiently covered in the publications by other authors.</p><p>In the implemented modification of the algorithm PQS boundary detector Kirsch was replaced by the boundary detector Canny. Further experiments were carried out according to the method of the ITU-R VT.500-13 (01/2012) using monochrome images treated with different types of filters (should be emphasized that an objective assessment of image quality PQS is applicable only to monochrome images). Images were obtained with a thermal imaging surveillance system. The experimental results proved the effectiveness of this modification.</p><p>In the specialized literature in the field of formal to evaluation methods pictures, this type of modification was not mentioned.</p><p>The method described in the publication can be applied to various practical implementations of digital image processing.</p><p>Advisability and effectiveness of using the modified method of PQS to assess the structural differences between the images are shown in the article and this will be used in solving the problems of identification and automatic control.</p>