Identification of Pests on Plants Using Clustering and Hybrid Approaches

Enormous agricultural yield is lost each year, because of quick pervasion by pest and insects. A great deal of research is being done worldwide to recognize logical procedures for early discovery/identification of these bio-aggressors. In the past years, a few methodologies dependent on computerization and digital image processing have become known to address this issue. The greater part of the calculations focus on pest identification and location, restricted to a greenhouse environment. Likewise, they include a few complex computations to accomplish the equivalent. In this paper, we developed a unique algorithmic approach to isolate and distinguish pest utilizing clustering and hybrid approaches. The proposed method includes decreased computational complexity and pest detection in green house environment. The whitefly, a bio-aggressor which represents a risk to a huge number of harvests, was picked as the pest of enthusiasm for this paper. The calculation was tried for a few whiteflies influencing various leaves and an accuracy of 96% of whitefly recognition was accomplished.