310 likes | 414 Views
Optimization of Object Extraction Based on One User-Prepared Sample. S. Rahnamayan, H.R. Tizhoosh, M.M.A. Salama University of Waterloo, Ontario, Canada MOPTA, Windsor, July 26, 2005. Agenda. Objective Proposed Approach Preliminary Results Comparison with Other Methods Conclusion
E N D
Optimization of Object Extraction Based on One User-Prepared Sample S. Rahnamayan, H.R. Tizhoosh, M.M.A. Salama University of Waterloo, Ontario, Canada MOPTA, Windsor, July 26, 2005
Agenda • Objective • Proposed Approach • Preliminary Results • Comparison with Other Methods • Conclusion • Future Works
Main Objective Acquisition of object extraction procedure from user-prepared sample(s) based on genetic optimization of morphological processing chains
Reasons for Developing Automated Image Processing Systems Dealing with huge number of images Saving experts valuable time Possibility of using in online applications Overcoming of inconsistent nature of human processing Supporting required high accuracy
Why Learning from a Small Number of Samples is Valuable ? Because : It reduces the expected level of expert participation which is the main obstacle for research and development. Preparing some manually generated samples to reflect the experts’ expectations is a reasonable requirement in all image processing environments.
Proposed Approach • Utilizing Mathematical morphology operations, as image processing tools, to build object extraction procedure • Using genetic algorithm, as optimizer tools, to find optimal parameters of above mentioned procedure
Morphological Operations They are shape-based operations Used to handle a wide range of image processing tasks, ranging from noise filtering to object extraction
Genetic Optimizer Optimal Ordering of Operations Parameter & Ordering Optimizer Procedure Applier Input Image Optimal Parameters Mathematical Morphology Operations Input Images Result images Gold Image Main Structure of Proposed Approach
Morphological Operations Chain as a Morphological Procedure • K3*{O(SE1)_C(SE2)}K1*E(SE3)K2*D(SE4) • 2. K1*E(SE3)K3*{O(SE1)_C(SE2)}K2*D(SE4) • 3. K1*E(SE3)K2*D(SE4)K3*{O(SE1)_C(SE2)} • 4. K3*{O(SE1)_C(SE2)}K2*D(SE4)K1*E(SE3) • 5. K3*{O(SE1)_C(SE2)}K2*D(SE4)K1*E(SE3) • 6. K3*{O(SE1)_C(SE2)}K2*D(SE4)K1*D(SE3) SE1, SE2, SE3, and SE4: Corresponding structural elements K1, K2, and K3 : Iteration times for operations O: Opening C: Closing D: Dilation E: Erosion
Start Genetic Optimization of MM Procedure Population Initialization Applying MM Procedure Computing of Dissimilarity Is Reached Ending Criteria? End Yes No Selection Crossover Mutation
Preliminary Results • Circle Extraction • Triangle Extraction • Rectangle Extraction • Object Extraction Applied for Grey-level Images
Utilized Measures Matching Index: Overall Matching Index:
Training for Object Extraction- Circle • Original image • Goal image • Generated image by MM procedure (94.48% similarity)
Training for Object Extraction- Triangle • Original image • Goal image • Generated image by MM procedure (85.01% similarity)
Training for Object Extraction- Rectangle • Original image • Goal image • Generated image by MM procedure (94.37% similarity)
Object Extraction Applied on Gray-scale Images • Grey scale image • Goal image • Generated image by MM procedure (76.77% similarity)
Some Results of Object Extraction in Grey Level Images 95.05% 96.05% 96.09% 96.71% Overall matching rate: 95.90% with standard deviation of 0.54% 95.63%
Training for Fully Rotation Invariant Triangle Extraction 1 1 Genetic Optimizer 2 2 Input Images Result Images 3 3 4 4 2 1 3 4 Gold Images
Comparing Proposed Approach with Knowledge-Based Learning Knowledge acquisition difficulties √ Unable of self-learning √ Difficult to avoid conflicts in large knowledge bases √ Knowledge reliability problem √ √ : Proposed approach solves it mostly or it is not applicable.
Comparing Proposed Approach with Sample-Based (NN) Learning Sample providing problem √ Problem of choosing the best architecture √ ~: Proposed approach solves it partially.
Conclusion The outstanding features of the proposed approach are as follows: - Training based on a few samples • Supporting (semi) automated image processing • Mostly invariant for noising, overlapping, translation, rotating, scaling, and duplicating.
Future Works - Extending functionality of the system to cover wider range of image processing tasks - Applying on medical image processing
Thank you for your attention and patience.