2020 IEEE International Conference for Innovation in Technology (INOCON) · 2020 · 39 citations · 26 references
ComputingComputer SciencePlant PathologyComputer VisionMachine LearningPlant Disease DiagnosisDeep LearningDisease DetectionDisease DiagnosisImage ClassificationPattern RecognitionTransfer LearningConvolutional Neural Network DetectionConvolutional Neural NetworksDeep Learning-based ClassificationDeep Learning-based DiagnosisDeep Learning-based Lesion ClassificationLeaf Blade MorphologyLeaf Shape AnalysisPlant Fungal Disease
Anthracnose is a plant disease caused by fungi that develop saucer shaped acervulus spore on leaves sunken, it lesions leaves and sooty mold mango fruit leaves. The Anthracnose disease affect the quality of mango fruit and productivity. Hence, it is essential to develop espousing model to detect lesion areas on Mango leaves, identify level of infection, diagnosis of the Anthracnose disease. The deep learning techniques are known for their performance and Convolutional Neural Networks(CNNs) are widely accepted model for pattern identification and image categorization. Therefore, this paper propose a modified version of VGGNet model called V2IncepNet that integrate best feature of VGGNet and Inception module. The VGGNet module extract basic feature while inception module perform extraction of high-dimensional features and classification of images. This paper use various features such as leaf color, venation, petiole condition, tip shape, tip conditions, leaf shape, leaf margin, dark spots on leaf blade and on midrib, margin of burns on leaf, leaf blade, midrib, and petiole. In our data set there are 2268 color images of Mango leaves, which include 1198 on-field self-captured real-time color images and 1070 color Mango leaves images downloaded from Plantvillage. The experiment result envisage that the proposed model can classify Anthracnose disease infection level on Mango leaves with accuracy not less than 92%. The proposed model is simple yet efficient.
26
Internet-of-Things (IoT)-Based Smart Agriculture: Toward Making the Fields Talk
Muhammad Ayaz, Zubair Sharif, Ali Mansour et al. · IEEE Access · 2019 · 1.2K citations · Full text
Engineering, Agricultural Engineering, Precision Agriculture +17