Those working in agricultural decision support systems (DSS) in agriculture are often disappointed that adoption by farmers of their work is not as high as they would hope. Perhaps farmers choose to work on the land because they don't like the kind of quantitative analysis that DSS provide? Are we trying to "fit a square peg in a round hole"? What makes agriculture different to other businesses where DSS have had a major impact? This paper reviews the use of DSS in agriculture and identifies some successful products and systems. From these examples we identify strategies for developing and delivering useful tools. We examine major trends that will shape future agriculture, and whether these trends will promote or discourage the use of DSS.

View full abstract

In order to develop a system for the automatic detection of leaf wilting, several methods of leaf region extraction were examined. Color images of bean, cucumber, and melon plants were obtained under artificial light in a growth chamber, under sunlight in the laboratory near a window, and under sunlight in a glasshouse, respectively. The optimum method for leaf region extraction seemed to depend on image acquisition conditions. For example, under artificial light, leaf region was satisfactorily extracted by using an intensity thresholding, whereas under sunlight in the laboratory, fairly good performance was obtained by using saturation thresholding or by G/B ratio thresholding.
When these optimal methods were applied for leaf region extraction, the y-coordinate of the center of gravity of the extracted area was used as an index of the leaf wilting.

View full abstract

Producers participating in farmers' markets themselves need to manage their articles, decide an amount and price of their commodity, and advertise their goods, which retailers have historically done. We developed several new applications to support farmers' markets using as a model a small farmers' market which opened in August 2001. Firstly, basic data of producers and their products were accumulated on a web server-side database (Web-DB), into which another system transmits current sales data from the marketcash register. Secondly, we developed some applications which process Web-DB to supply necessary information to retailers, producers and customers through the Internet. For instance, this system supports the following decision-making: Producers themselves decide retail prices and shipping amounts judging from their stock and sales, and customers decide whether they will drop in at a farmers' market according to some information on the farmers' market's selling prices and goods while shopping at a big store close to the farmer's market. Since the system utilizes the web browser as its interface, it can also be used in the fields or outside with a cellular phone. So it can help both producers and customers make decisions because they can get useful information everywhere with their cellular phones. In addition, an application to record the attendance of producers at the morkets is descried. Only three months have passed since this system was made available, but we discussed some results that occurred: How the producers have used it and which improvements are needed.

View full abstract

In March 2000, we first developed an integrated database system called "PaDB (Pathology Database)" for the purpose of supporting preventive work for crop diseases and pests. The system is composed of three separate databases --Plant protection guidebook, Illustrated book of crop diseases and pests, and Agricultural chemicals catalog-- and a retrieval interface system that integrates these databases. The purpose of this research is to establish the practical use of the PaDB system. More specifically, we performed experiments in: (1) distributing databases; (2) adding a database of cultivation calendars; and (3) developing a retrieval interface that can be used with an i-mode cell phone. As a result, it became feasible be use the PaDB to: (1) link distributed databases that have their own individualities together; (2) supporting preventive measures for crop diseases and pests; (3) retrieval of information on the farm.

View full abstract

This paper describes development of a rule-based expert system for the diagnosis of main diseases (like powdery mildew, anthracnose, black spot etc, their causing agent and remedy) of Indian mango tree. The objective is to provide computer-based support for agricultural specialist or farmer in fruiticulture. The proposed expert system makes diagnosis on the basis of response/responses of the user made against queries related to particular disease symptoms. The system has been developed in Borland's Turbo-prolog 2.0 environment. The knowledge base of the system contains knowledge about symptoms of 14 diseases of Indian mango tree appearing during fruiting season or in non-fruiting season. The result given by the system has been found to be sound and consistent.

View full abstract