Identification of the Green Leafhopper (GLH) on a light trap using digital image processing

Abstract

Leafhopper, any of the small, slender, often beautifully coloured and marked sap-sucking insects of the large family Cicadellidae of the order Homoptera. They are found on almost all types of plants. However, individual species are host-specific. Although a single lea fhopper does no damage to a plant, collectively they can be serious economic pests. Their fee ding may injure the plant in any of several ways by removing sap, destroying chlorophyll, transmitting diseases, or curling leaves. Green Leafhopper (GLH) Nephotettix virescens can spread tungro virus to rice plants. Both nymphs and adults feed by extracting plant sap with their needle-shaped mouthparts. The first symptoms of infestation are yellow, transparent spots that appear mainly on the tips and along the mid-ribs of the paddy leaves. These are soon followed by grayish-white, and later brown spots on the leaves and leaf sheaths of the young plants. Knowing the population number of GLH in paddy field is important to avoid damage to the plant. Early prevention techniques such as insecticides spraying can be taken to prevent the increase of GLH population. However, the traditional method of counting GLH population took hours to finish. In this research, the use of machine vision and digital image processing will helps to identify and shorten the time period of counting GLH population. This will helps farmers to deliver the fastest prevention method to their paddy field and could avoid damage to plant. An algorithm is required for this new method. The result will be compared to the actual result counted by the experience worker to validate the accuracy of the algorithm.