Real-Time Object Measurement and Classification

This book contains papers presented at the NATO Advanced Research Workshop on "Real-time Object and Environment Measurement and Classification" held in Hotel Villa del Mare, Maratea, Italy, August 31 - September 3, 1987. This workshop was organized under the NATO Special Programme on Sensory Systems for Robotic Control. Professor Eric Backer, Delft University of Technology, The Netherlands and Professor Erdal Panayirci, Technical University of Istanbul, Turkey were the members of the organizing committee for this workshop. There were four major themes of this workshop: Real-time Requirements, Feature Measurement, Object Representation and Recognition, and Architecture for Measurement and Classification. A total of twenty-five technical presentations were made. These talks covered a wide spectrum of topics including hardware implementation of specific vision algorithms, a complete vision system for object tracking and inspection, using three cameras (trinocular stereo) for feature measurement, neural network for object recognition, integration of CAD (Computer-Aided Design) and vision systems, and the use of pyramid architectures for solving varioos computer vision problems.

1. Two real-time architectures for image processing and computer vision.- 2. A flexible and intelligent system for fast measurements in binary images for in-line robotic control.- 3. Object recognition and real-time relative state estimation under egomotion.- 4. Hardware-software trade-offs in robot vision.- 5. Hardware versus software implementations of fast image processing algorithms.- 6. Robot perception systems: some design issue.- 7. Two parallel algorithms for the analysis of random images.- 8. Shape detection using the adaptive Hough transform.- 9. Shape analysis based on boundary curve segmentation.- 10. Progress in trinocular stereo.- 11. Feature extraction using structured light.- 12. A clustering approach to texture classification.- 13. Spatial feature measurements in parallel structures.- 14. What an autonomous robot can learn about its environment starting from scratch using image sequences.- 15. Three-dimensional object recognition based on multiview representation and extended Gaussian image.- 16. Symbolic surface descriptors.- 17. A novel architecture for real time pick up of 3D motion and 3D layout information from the flow of the optic array.- 18. Object identification and automatic learning using CA vision, CAD and AI based approach.- 19. Real-time restoration and segmentation algorithms for hidden Markov mesh random fields image models.- 20. Coherent architectures for machine vision.- 21. Architectural solutions for intermediate-level vision.- 22. Statistical optimization of image processing architectures for maximum resource utilization and throughput.- 23.Object recognition in multi-resolution systems.- 24. Multisensor knowledge systems.- 25. Knowledge representation for three-dimensional sensor fusion with context truth maintenance.- List Of Participants.