Application of machine vision to human shape analysis techniques in leaf and plant identification: An intelligent vision structure
Abstract
Intelligent vision systems capable of automatically interpreting images are the next generation in machine vision. These systems will be functional to nonexperts in the machine vision field, and will be flexible and reprogrammable for general use. Addition of intelligence into vision systems requires an understanding and structuring of human visual and scene interpretation techniques. This research was designed to investigate the interrelationships and combine the areas of image processing, feature extraction and techniques of human scene interpretation. The research is especially unique in its rule-based structuring of subjective, qualitative human-level shape features and quantitative machine-level shape features. Additionally a unique approach was implemented to extract plant/leaf shape features using information gathered from critical points along object borders. The critical points were determined using the object's centroid. The image processing portion of the program followed a procedural format with examples of images processed by various routines displayed on a screen for user evaluation, which allowed the system to act as a programming aid.
Degree
Ph.D.
Advisors
Miles, Purdue University.
Subject Area
Agricultural engineering|Artificial intelligence
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.