Weeds are a major constraint to the success of crop growth. Crops worldwide are affected by weeds which reduces yields. Even more challenging is how long farmers usually take to determine how they have infested their farms. The traditional method of detecting the weeds by farmers is by use of their human eyes. This has been found to be tedious in a way that is time consuming and labour consuming, especially when dealing with a large piece of land. Also, this method only requires physical presence of the farmers. The objective of this project therefore was to develop an AI weed detection and monitoring system (mobile application) that acts as a companion to farmers by providing them with day-today information about the infestation of weeds on their farms to enable them take immediate, knowledgeable and appropriate decisions. The project involves the system taking pictures of the farm and then processing them (by the trained model), from which the corresponding answers of whether or not, and which type of weed or crop detected is provided. Various methods were used to gather requirements which include interviews and study of the existing systems. In conclusion, the use of this AI weed detection and monitoring system enables farmers and agriculturalists to take effective and informative decisions about weed control which increases their rate of crop production. In addition, it facilitates remote monitoring of the weed growth on the farm by the farmer effectively.

show more