EVALUATION OF MEASUREMENTS OF THE DISTANCE TO THE OBJECT IN THE STUDY OF ITS GRAPHIC IMAGE
Pages 54-65
An important element of modern automated systems of management, monitoring and control of remote access modules is information about the state and behavior of a static or moving object. In many existing and planned monitoring systems processing graphical image of the object is used, which is obtained by the photo detectors and, thus, the possibility of determining geometric and kinematic parameters of a moving object is significantly reduced due to various aspects of image acquisition, one of such aspects is blur. In the present work, algorithms of the primary information processing obtained on the basis of the graphic image study of a movable or stationary object are improved using the methods and procedures of statistical analysis that allow approximating theoretical results to experimental results. The use of statistical analysis and probabilistic approach increase the accuracy of the determined characteristics, applicability of calculation procedures of the state parameters (size, shape, distance from the observer) and behavior of the object (speed and direction) and reduce the computational complexity of the final algorithm. The Bayesian estimation was obtained based on the use of quadratic, rectangular and simple loss function under normal, Laplace, uniform and lognormal distribution of errors, which allow drawing conclusions about the intervals of various models and algorithms to determine the parameters of different objects. When using the statistical approach it is taken into account that the errors are random in nature and may be considered a variety of probability density functions (normal, lognormal, Laplace and uniform distributions to minimize risks under different loss functions (quadratic, rectangular, linear), and then evaluated using the method of least squares, method of least modules and the Bayesian approach. When performing the evaluation the properties of unbiased, consistency and efficiency are important. Sustainable procedure should have the following properties: for the selected model, the procedure should be close to optimum efficiency; the results should be close to nominal, calculated for the adopted model; the effect of large errors must be eliminated. We use the minimax method of Huber, which assumes that the best estimate will not be worse than in the case of the “least favorable” density distribution. Decision rule is based on the definition of such density that minimizes the information of Fischer that is the variance function of the contribution of the sample. In the present study we offer the procedure of finding a theoretical function based on the assumption that the errors are subjected to the known laws of distribution: normal (Gaussian), Laplacian, uniform, lognormal. This is a significant advantage of the proposed methodology compared to the one used in the previous works of the authors, it is proposed here to use the Bayesian estimation of the measurements as unknown theoretical function that needs to be obtained closer to the observational measurements.
DOI: 10.22227/1997-0935.2015.10.54-65
- Sun Z., Bebis G., Miller R. On-road Vehicle Detection Using Optical Sensors: A Review. Proceeding of the IEEE International Conference on Intelligent Transportation Systems. 2004, vol. 6, pp. 125—137.
- Nayar S.K., Nakagawa Y. Shape from Focus: An Effective Approach for Rough Surfaces. Proceeding CRA90. 1990, vol. 2, pp. 218—225. DOI: http://dx.doi.org/10.1109/ROBOT.1990.125976.
- Rabe C., Volmer C., Franke U. Kalman Filter Based Detection of Obstacles and Lane Boundary. Autonome Mobile Systeme. 2005, vol. 19, pp. 51—57. DOI: http://dx.doi.org/10.1007/3-540-30292-1_7.
- Loktev D.A., Loktev A.A. Determination of Object Location by Analyzing the Image Blur. Contemporary Engineering Sciences. 2015, vol. 8, no. 11, pp. 467—475. DOI: http://dx.doi.org/10.12988/ces.2015.52198.
- Rajagopalan A.N., Chaudhuri S. An MRF Model-Based Approach to Simultaneous Recovery of Depth and Restoration from Defocused Images. Transactions on Pattern Analysis and Machine Intelligence. 1999, vol. 21, no. 7, pp. 577—589. DOI: http://dx.doi.org/10.1109/34.777369.
- Gaspar T., Oliveira P. New Dynamic Estimation of Depth from Focus in Active Vision Systems. Preprints of the 18th IFAC World Congress Milano (Italy) August 28 — September 2. 2011, pp. 484—491. DOI: http://dx.doi.org/10.5220/0003356904840491.
- Lelegard L., Vallet B., Bredif M. Multiscale Haar Transform for Blur Estimation from a Set of Images. International Archives of Photogrammetry : Remote Sensing and Spatial Information Science. Munich, Germany, October 5—7, 2011, pp. 65—70.
- Lin H.-Y., Chang C.-H. Depth from Motion and Defocus Blur. Optical Engineering. December 2006, vol. 45 (12), no. 127201, pp. 1—12. DOI: http://dx.doi.org/10.1117/1.2403851.
- Levin A., Fergus R., Durand Fr., Freeman W.T. Image and Depth from a Conventional Camera with a Coded Aperture. ACM Transactions on Graphics. 2007, vol. 26, no. 3, article 70, pp. 124—132.
- Loktev A.A., Loktev D.A. Metod opredeleniya rasstoyaniya do ob”ekta putem analiza razmytiya ego izobrazheniya [Method of Determining the Distance to the Object by Analyzing its Image Blur]. Vestnik MGSU [Proceedings of Moscow State University of Civil Engineering]. 2015, no. 6, pp. 140—151. (In Russian)
- Sizikov V.S., Rimskikh M.V., Mirdzhamolov R.K. Reconstructing Blurred Noisy Images Without Using Boundary Conditions. Journal of Optical Technology. 2009, vol. 76, no. 5, pp. 279—285. DOI: http://dx.doi.org/10.1364/JOT.76.000279.
- Elder J.H., Zucker S.W. Local Scale Control for Edge Detection and Blur Estimation. IEEE Transaction on Pattern Analysis and Machine Intelligence. 1998, vol. 20, no. 7, pp. 699—716. DOI: http://dx.doi.org/10.1109/34.689301.
- Alfimtsev A.N., Loktev D.A., Loktev A.A. Razrabotka pol’zovatel’skogo interfeysa kompleksnoy sistemy videomonitoringa [Development of a User Interface for an Integrated System of Video Monitoring]. Vestnik MGSU [Proceedings of Moscow State University of Civil Engineering]. 2012, no. 11, pp. 242—252. (In Russian)
- Alfimtsev A.N., Loktev D.A., Loktev A.A. Sravnenie metodologiy razrabotki sistem intellektual’nogo vzaimodeystviya [Comparison of Development Methodologies for Systems of Intellectual Interaction]. Vestnik MGSU [Proceedings of Moscow State University of Civil Engineering]. 2013, no. 5, pp. 200—208. (In Russian)
- Jiwani M.A., Dandare S.N. Single Image Fog Removal Using Depth Estimation Based on Blur Estimation. International Journal of Scientific and Research Publications. 2013, vol. 3, no. 6, pp. 1—6.
- Loktev A.A., Alfimtsev A.N., Loktev D.A. Algoritm raspoznavaniya ob”ektov [Algorithm of Object Recognition]. Vestnik MGSU [Proceedings of Moscow State University of Civil Engineering]. 2012, no. 5, pp. 194—201. (In Russian)
- Robinson Ph., Roodt Yu., Nel A. Gaussian Blur Identification Using Scale-Space Theory. Faculty of Engineering and Built Environment. University of Johannesburg, South Africa, 2007, pp. 68—73.
- Langley P. User Modeling in Adaptive Interfaces. Proc. of the Seventh Intern. Conf. on User Modeling. 1997, pp. 357—370.
- Trifonov A.P., Korchagin Yu.E., Trifonov M.V., Chernoyarov O.V., Artemenko A.A. Amplitude Estimate of the Radio Signal with Unknown Duration and Initial Phase. Applied Mathematical Sciences. 2014, vol. 8, no. 111, pp. 5517—5528. DOI: http://dx.doi.org/10.12988/ ams.2014.47588.
- Chernoyarov O.V., Sai Si Thu Min, Salnikova A.V., Shakhtarin B.I., Artemenko A.A.Application of the Local Markov Approximation Method for the Analysis of Information Processes Processing Algorithms with Unknown Discontinuous Parameters. Applied Mathematical Sciences. 2014, vol. 8, no. 90, pp. 4469—4496. DOI: http://dx.doi.org/10.12988/ ams.2014.46415.