Insect Identification in Field Crops Using CNN

In recent times, many individuals have been leaving the agricultural sector due to challenging environmental conditions and meager income prospects. However, there is a growing interest among the younger generation in pursuing farming, despite their limited knowledge in the field. To support this resurgence in agriculture, an innovative application and website have been developed for insect identification and classification. This technology aims to empower Indian farmers who struggle with detecting pests that harm paddy crops, a significant contributor to crop damage and yield loss. By utilizing shape features and machine learning techniques, a proposed algorithm combines foreground extraction, contour identification, and a CNN model to effectively identify pests like Tissueborers, Gundhi bugs, and Paddy Gall flies within the intricate backgrounds of paddy fields. This pest detection method offers faster computation times and high accuracy, enabling the early identification of insects during paddy crop growth. Ultimately, by preventing insect-related damage, this approach has the potential to boost both crop yield and quality, leading to enhanced agricultural productivity.