|
|
Weed Identification Using a Picture Based Hierarchical Classification System
|
|
|
Abstract: The goal of our work is to develop a system that will address all aspects of irrigated wheat management in Egypt including pest identification and remediation. In this paper, we present our implementation of an expert system for weed identification. The approach we take to solve the problem of weed identification is based upon the Generic Task Approach to expert systems development pioneered by Chandrasekaran et al. (Chandrasekaran, 1986) and two approaches to weed identification outlined by Hanf (Hanf, 1990) and Behrendt and Hanf (Behrendt & Hanf, 1979). We have identified weed identification as a problem. Therefore, Hierarchical Classification (Gomez & Chandrasekaran, 1981) is the individual GT that provides the problem solving template for our weed identification. The computer-based approach enabled us to overcome several shortcomings of traditional, dichotomous keys. Through the use of pictures, we largely avoided technical terms. Furthermore, the system allows for multiple decisions at each level. The system also enables the user to backtrack and revise decisions.
|
| URL |
|
| Publication year |
1994
|
| Organization Name |
|
| Country |
United States
|
| City |
Seattle, Washington
|
| serial title |
Agriculture and Natural Resource Management WorkshopAAAI-94
|
| Author(s) from ARC |
|
| External authors (outside ARC) |
|
| Agris Categories |
Documentation and information
|
| Publication Type |
Conference/Workshop
|