AN IMAGE-ANALYSIS SYSTEM DEVELOPED FOR EVALUATION OF COLEOMEGILLA-MACULATA LARVAES BEHAVIOR

A method based on machine vision was developed to replace manual obser vations required in a study of prey discrimination of the twelvespotte d ladybird beetle (Coleomegilla maculata) larvae. The system recorded the movement of the II or IV instar larva while the beetle was in cont act with a group of 36 cabbage looper (Trichoplusia iii) eggs, half of which had been parasitized by Trichogramma evanescens before the test . Since the beetle larvae are approximately 20 to 35 times larger than the eggs, the system was programmed to locate the eggs and to follow the movement of the larval head. A three step procedure was developed to detect the larval head position with less than 0.1% of error. Resul ts recorded by the system were more objective than those obtained visu ally. The developed method and different parameters used within this m ethod are presented in this paper.