Expert system for identification of weed seedlings

Abstract

Weeds can be controlled effectively only when the management practices are employed at and early growth stage. Therefore, identifying weeds at seedlings stage is very essential for designing an efficient strategy for their management. However, it is a difficult task unless one has enough knowledge and experience. To help identifying weeds at seedling stage, an expert system has been developed at the Directorate. The expert system uses scientific name-based classification and a mix of the text description and images. The system is supported by a database containing information on 138 weed species with colour images of weed seedling at five growth stages. The expert system was evaluated following the conventional expert system evaluation methodologies. Results indicated that non-expert users were able to make weed seedling identification using the expert system more efficiently.