Automatic vision programming

Abstract Techniques for automatically generating optimal vision programs from high-level task descriptions are presented. Vision programs are the object models that describe strategies to recognize and locate objects in an image. The effectiveness of the program depends on the features used for recognition and the order in which the features are evaluated. We describe three probabilistic feature utility measures and a cost function based on program execution time that serve as the basis of our technique. Techniques for computing the utility measures and combining them to generate the best vision program are described and are demonstrated on a realistic vision application.

[1]  W. Eric L. Grimson,et al.  Localizing Overlapping Parts by Searching the Interpretation Tree , 1987, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[2]  Ramesh C. Jain,et al.  Recognizing partially visible objects using feature indexed hypotheses , 1986, IEEE J. Robotics Autom..

[3]  Alex Pentland,et al.  Perceptual Organization and the Representation of Natural Form , 1986, Artif. Intell..

[4]  D. W. Thompson,et al.  Three-dimensional model matching from an unconstrained viewpoint , 1987, Proceedings. 1987 IEEE International Conference on Robotics and Automation.

[5]  Chris Goad,et al.  Special purpose automatic programming for 3D model-based vision , 1987 .

[6]  Thomas C. Henderson,et al.  CAGD-Based Computer Vision , 1989, IEEE Trans. Pattern Anal. Mach. Intell..

[7]  Takeo Kanade,et al.  Automatic generation of object recognition programs , 1988, Proc. IEEE.

[8]  George C. Stockman,et al.  Object recognition and localization via pose clustering , 1987, Comput. Vis. Graph. Image Process..

[9]  Yehezkel Lamdan,et al.  Geometric Hashing: A General And Efficient Model-based Recognition Scheme , 1988, [1988 Proceedings] Second International Conference on Computer Vision.