Identification Spectral Signature of Weed Species in Rice using Spectroradiometer Handheld Sensor

Abstract

Productions of rice in Malaysia is less compared to other rice producers such as Thailand. One of the reasons is lack of management in weeds. Losses due to weeds must be solved in rice production system, they interfere with the activities involved in the field throughout crop growing period. The main problem in rice field is to detect the presence of weeds among the paddy plant. It is hard to detect from plain sight for weeds in the green field by farmers. This research developed a technique to detect and control the weeds in the rice field. Spectral signature of weeds was collected and stored in the mobile apps for weed management application. The spectral signature of weeds species is unique and has their own characteristics. It can be used as a reference on how to detect weeds in the rice field using remote sensing. In this study, the result outcomes are mainly targeting to help all the paddy farmers in Malaysia to increase their production numbers of paddy. In addition, paddy farmers will be introduced with much advance technique to spot on the weeds that have been infesting in their paddy field and can prepare with what strategy to control on different kind of weeds. Different weed species has different physiological characteristics and it requires different type of strategy to counter the growth in the paddy field. Every weed in the paddy field need to be recognized in order to plan a control strategy. By using only naked eyes of a human and the limited view through the side view of the paddy field contributes to the problem of recognizing type of weeds that infesting the area.