PROSPECTIVE TECHNOLOGIES OF DIGITAL IMAGE PROCESSING IN CONTROL OF OBJECT SURFACESROCKET AND SPACE EQUIPMENT
State control method of visual inaccessible surfaces of technical objects was examined by research of digital images which contains information about their state and quality. Received data lets classity compare and mark out the classes of normal and abnormal digital images. Results and conclusion of visual analysis were confirmed by statistical treatment of matrix measurements of digital images. In the context of a lack of a priori data on the in formativeness and statistical regularities of experimental matrices for measuring the brightness of digital images of technical objects being developed and inaccessible for observation, their condition and quality should be monitored in a comprehensive manner and in stages. At the first stage, by visual analysis of digital images, their classification is carried out with the allocation of a class of anomalous digital images and a class of images, the controlled surfaces of which are considered normal. At the second stage, by means of statistical processing of measurement matrices, data are generated for visual - analytical analysis and verification of hypotheses about the state of the controlled surfaces of technical objects and confirmation of the results of their classification by visual examination of digital images. Methods of processing matrices of such experimental measurements are an estimation of their mathematical expectations, variances, correlation coefficients, empirical functions and probability distribution laws. These integrated estimates are informative indicators of statistical homogeneity of measurement samples as random variables. They are used in the design, development and testing of rocket and space technology facilities. As a result of the work, a new method of visual-analytical processing of digital images was proposed to control the surfaces of objects of rocket and space technology during their design and testing.
В.Т. Фесенко, Т.Ю. Фесенко. Компьютерная обработка и распознавание изображений: учеб. пособие. СПб: СПбГУ ИТМО, 2008. 192 с.
Т.А. Манько, И.А. Гусарова, К.В. Козис. Контроль состояния визуально недоступных поверхностей технических объектов. Системные технологии: сб. науч. тр, Днепр, 2017. Вып.2(109), С.87-94.
Malik Jitendra, Sergey Belogie, Txomas Leung, Jianbo Shi. Contour and Texture Analysis for Image Segmentation. International Journal of Computer Vision, 2001. V. 43. No. 1. P. 7-27.