Using AI Planning Techniques to Automatically Generate Image Processing Procedures

This paper describes work on the Multimission VICARPlanner (MVP) system to automatically construct executable imageprocessing procedures for custom image processing requests for the JPL Multimission Image Processing Lab (MIPL). This paper focuses on twoissues. First, large search spaces caused by complex plans required the use of hand encodedcontrol information. In order to address this in a mannersimilar to that used by humanexperts, MVPuses a decomposition-based planner to implementhierarchical/skeletal planning at the higher level and then uses a classical operator based planner to solve subproblemsin contexts defined by the highlevel decomposition. Second, the image processing domainis characterized by large amountsof search to find the correct program options for images (e.g. operator effects), rather than search amongdifferent programs(e.g. planning operators) and manyof these program options are incompatible (i.e. certain combinations cannot be used). MVPrepresents these intera...

• Type: Conference
• Year: 1994
• Where: AIPS
• Authors: Steve A. Chien
• Keywords: AIPS 1994, Artificial Intelligence, Image Processing, Multimission Image Processing, Program Option