Combination of image processing and artificial neural networks as a novel approach for the identification of Bemisia tabaci and Frankliniella occidentalis on sticky traps in greenhouse agriculture

Abstract

Integrated Pest Management (IPM) lies at the core of the current efforts to reduce the use of deleterious chemicals in greenhouse agriculture. IPM strategies rely on the early detection and continuous monitoring of pest populations, a critical task that is not only time-consuming but also highly dependent on human judgement and therefore prone to error. In this study, we propose a novel approach for the detection and monitoring of adult-stage whitefly (Bemisia tabaci) and thrip (Frankliniella occidentalis) in greenhouses based on the combination of an image-processing algorithm and artificial neural networks. Digital images of sticky traps were obtained via an image-acquisition system. Detection of the objects in the images, segmentation, and morphological and color property estimation was performed by an image-processing algorithm for each of the detected objects. Finally, classification was achieved by means of a feed-forward multi-layer artificial neural network. The proposed whitefly identification algorithm achieved high precision (0.96), recall (0.95) and F-measure (0.95) values, whereas the thrip identification algorithm obtained similar precision (0.92), recall (0.96) and F-measure (0.94) values.