ESTIMATION OF QUALITY METHODS DISGUISE IMAGES FOR DETECTION EDGE CONTOURS
DOI:
https://doi.org/10.18372/2310-5461.18.4936Keywords:
image of masking (disguise), GT – image, method of assessment, measure estimates, quality of detection, quality of the localizationAbstract
The article is dedicated to quality assessment methods for semantics of images is masking images (search, detection and localization of the contour structure in the processed image). The purpose of this paper is the analysis of ways to measure the quality of image masking methods (quality of detection and localization of circuits) for practical realization in automatic image processing and video data to improve the quality of treatment. The article analyzes well-known the measure for evaluation the quality of masking, which are divided into two groups of measures: assessing the quality of the detection and evaluation of the localization of the contours. The main advantages and disadvantages of measures assessing the quality of masking methods. Based on the synthesis of practical application of masking methods in the known systems image processing and experimental data the author comes to the conclusion that the proposed Methods of evaluation the quality of masking images can give a sufficiently detailed quantitative characterization of concealment methods. Displaying approaches to comparative estimation methods of masking to select a specific method (methods) for the purpose of practical realization in image processing systems (video data). Finally it was concluded that the justification for the choice of one of the existing methods of masking images to use of the comparative quantitative measures of the quality masking images with ground truth images (images that contain a common understanding of the researcher border).
References
Gonsales R. S. Digital image processing / R. S. Gonsales, R. E. Vuds. — М. : the Technosphere, 2006. — 1072 with.
Прэтт U. K. Tsifrovaja obrabotka izobrazheny / U. K. Прэтт. — М. : the World, 1982. — 792 with.
Ablamejko S. V. Image processing: technology, methods, application / S. V. Ablamejko, D. M. La gunovsky. — Minsk : Amalfeja, 2000. — 303 with.
Білинський Я. Я. Meтоди обробки зображень в комп’ютеризованих оптико-електронних системах / Я. Я. Білинський. — Вінниця : ВНТУ, 2010. — 272 with.
Herskovits A., Binford T. On Boundary Detection, MIT Project MAC, Artificial Intelligence Memo 183, July 1970.
Fram J. R., Deutch E. S. On the Evalution of Edge Detection Schemes and Their Comparison with Human Performance, IEEE Trans. Computers, C-24, Vol. 6, p. 616– 628 (June 1975).
Bowyer K., Kranenburg A., Dougherty S. Edge detector evaluation using empirical ROC curves, Computer vision and Image Understanding. 2001. Vol. 84. № 1, p. 77–103.
Gribkov I. V. [et al.]. Edge Detection under Affine Transformations: Comparative Study by PICASSO 2 System, WSEAS Transactions on Signal Processing. 2006. Is. 9. Vol. 2, p. 1215–1221.
Prieto M. S., Allen A. R. A similarity metric for edge images, IEEE Trans. Pattern Anal. Mach. Intell. 2003. Vol. 25. № 10, p. 1265–1273.
Osipov A. A fuzzy approach to performance evaluation of edge detectors, in Lecture Notes in Signal Science, Internet and Education, WSEAS Press / A. Osipov. — 2007, p. 94–99.
Some's fungi questions of the quantitative estimation of productivity of detectors of boundaries / I. V. Gribkov, A. V. Zaharov, P. P. Koltsov, etc. // Software solutions and systems. — № 4, 2011, with. 13–19.
Wang Z. Image quality assessment: From error visibility to structural similarity. IEEE transaction on Image Processing / Z. Wang, A. Bovik, H. Sheikh. — 2004. — Vol. 13, 4., p. 309–312.
A. J. Baddeley. Errors in binary images and Lp version of the Hausdorff Metric, Nieuw Archief voor Wiskunde / A. J. Baddeley. 1992. Vol. 10, pp.
–183.
