Strawberry Leaf and Fruit Pest Detection using Deep Learning

The project proposes an image pattern classification to identify adulteration in strawberry with a combination of texture and color feature extraction. The purpose of this research is to find appropriate features that can identify strawberry disease. Firstly, normal and adulterated images are collected and processed. Then, features of shape, color and texture are extracted from these images. After that, these images are classified by the support vector machine classifier. A combination of several features is used to evaluate the appropriate features to find distinctive features for identification of adulteration. This method uses YOLOv3 which is real time object detection algorithm that identifies specific objects in videos, live feeds, or images, it uses features learned by a deep convolutional neural network to detect an object where YOLOv3 is an improved version of YOLO and YOLOv2 and this model is known as DARKNET-53 which has 53 convolution layers with residual connections. Here the last three layers of draknet-53 are ignored as these layers are mainly used for image classification and we are using draknet-53 only to extract image features so these layers will not be needed. YOLO is implemented using Keras and OpenCV deep learning libraries and is combined with Convolution Neural network algorithm method VGG16 which has accuracy of 92.7% top5 test on ImageNet dataset which contains 14 million images belonging to 1000 classes. VGG16 is the best performing architecture in ILSVRC challenge in 2014.It was a runner up in classification task with top-5 classification error 7.32%.