Identification of Brown planthopper on a light trap using digital image processing

Abstract

The high population density of brown planthopper (BPH) in paddy fields can cause paddy trees to turn yellowish-brown before becoming war and dry (hopperburn) then kill the plant. Pest damage caused by this population may result in high losses to farmers. Therefore, in order to prevent this BPH population from spreading further, the A3-sized light trap is used to monitor the population of war at one time. Calculation process is done manually by an experienced worker but it takes almost six hours to complete a light trap. So, another calculation method needs to be done to expedite the calculation process of this war. In this study, machine vision and image processing techniques are used to identify and differentiate the characteristics of BPH from other insects. In conclusion, all of the three objectives in this study have been achieved and the results of precision algorithms used in this study have obtained an accuracy of 90% after compared to actual results carried out by MARDI Seberang Perai.