A genetic algorithm for the identification and segmentation of known motion-blurred objects

This paper presents a Genetic Algorithm based technique capable of identifying moving objects whose image is blurred due to fast relative motion with respect to the acquisition camera. Moreover, also the extension of the motion and the rotation of the object during the acquisition time can be accurately inferred. The proposed method is applicable when the geometric properties of the object were previously recorded in a database. Extensive testing shows that the proposed algorithm yields high success rates of correct identification of both the bottle species and of its motion with a limited number of chromosomes. The computing time is reasonably fast and the algorithm can be applied in real-time applications.

[1]  Sebti Foufou,et al.  A multiagent system approach for image segmentation using genetic algorithms and extremal optimization heuristics , 2006, Pattern Recognit. Lett..

[2]  Anil Mital,et al.  Manual, Hybrid and Automated Inspection Literature and Current Research , 1993 .

[3]  William T. Freeman,et al.  Removing camera shake from a single photograph , 2006, SIGGRAPH 2006.

[4]  Dzuraidah Abd. Wahab,et al.  Development of a Prototype Automated Sorting System for Plastic Recycling , 2006 .

[5]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .

[6]  Anil Mital,et al.  A comparison between manual and hybrid methods in parts inspection , 1998 .

[7]  Hassan Basri,et al.  Application of automated image analysis to the identification and extraction of recyclable plastic bottles , 2009 .

[8]  Klaus Diepold,et al.  Comparison of motion de-blur algorithms and real world deployment , 2006 .

[9]  Jian-Feng Cai,et al.  Blind motion deblurring using multiple images , 2009, J. Comput. Phys..

[10]  Cristián Zegers Ariztía,et al.  Manual , 2002 .

[11]  Linda G. Shapiro,et al.  Image Segmentation Techniques , 1984, Other Conferences.

[12]  V. B. Surya Prasath,et al.  Ringing Artifact Reduction in Blind Image Deblurring and Denoising Problems by Regularization Methods , 2009, 2009 Seventh International Conference on Advances in Pattern Recognition.

[13]  Hassan Basri,et al.  An Efficient Segmentation Technique for Known Touching Objects Using a Genetic Algorithm Approach , 2007, Australian Conference on Artificial Intelligence.

[14]  Daniela Iacoviello,et al.  Filtering image sequences from a moving object and the edge detection problem , 2006, Comput. Math. Appl..