Outdoor field machine vision identification of tomato seedlings for automated weed control

A machine vision system to detect and locate tomato seedlings and weed planes in a commercial agricultural environment was developed and tested. Images acquired in agricultural outdoor tomato fields under natural-light-only conditions were studied extensively, and an environmentally adaptive image segmentation algorithm was developed to improve machine recognition of plants under these conditions. To overcome the plant leaf occlusion problem, an object partition algorithm was used to separate the overlapped leaves. Four morphological features were used in the plant leaf classification, and several structural features were used in a syntactic procedure to identify the whole tomato plant and its stem location in the field. The system was able to identify the majority of non-occluded target plant cotyledons, and to locate plant centers even when the plant was partially occluded. Of all the individual target crop plants, 65% to 78% were correctly identified and less than 5% of the weeds were incorrectly identified as crop plants.