Knowledge based system for weed seed identification

Abstract

The term ‘weed’ always has a negative connotation not only by its presence as a plant but as a seed also. The movement of most of the weeds from one location to the other is mainly through their seeds. Correct identification of weed seeds is therefore necessary for strict quarantine to check the spread of weed seeds from one place to another. Accurate identification of weed seeds requires skill and good judgment on the part of the examiner and it is a difficult task for a layman. Therefore, availability of a state of art technology for identification of weed seeds is very much needed. For identification of weed seeds, a Knowledge Based System (KBS) contains information about 120 weed seeds with 11 parameters each has been developed at the Directorate of Weed Research, Jabalpur (India) using Visual Basic. Net as front-end application and MS Access as back-end application with user-friendly menus. In this study, a rule based system for identification of weed seeds was developed that helps the stakeholder in identifying the weed seeds.