Precision agriculture is a rapidly developing field aimed at addressing current concerns about agricultural sustainability. Machine learning is the cutting edge technology underpinning precision agriculture, enabling the development of advanced disease detection and classification methods. This paper presents a review of the application of machine learning and deep learning techniques in precision agriculture, specifically for detecting and classifying plant diseases. We propose a novel classification scheme that categorizes all relevant works in the associated classes. We separate the studies into two main categories depending on the methodology that they use (i.e., classification or object detection). In addition, we present the available datasets for plant disease detection and classification. Finally, we perform an extensive computational study on five state-of-the-art object detection algorithms on PlantDoc dataset to detect diseases present on the leaves, and eighteen state-of-the-art classification algorithms on PlantDoc dataset to predict whether or not there is a disease in a leaf. Computational results show that object detection accuracy is high with YOLOv5. For the image classification task, the networks ResNet50 and MobileNetv2 have the most optimal trade-off on accuracy and training time.

In recent years, the production of municipal solid waste has constantly been increasing. Recycling is becoming more and more important, as it is the only way that we can have a clean and sustainable environment. Recycling, however, is a process that is not fully automated; large volumes of waste materials need to be processed manually. New and novel techniques have to be implemented in order to manage the increased volume of waste materials at recycling factories. In this paper, we propose a novel methodology that can identify common waste materials as they are being processed on a moving belt in waste collection facilities. An efficient waste material detection and classification system is proposed, which can be used in real integrated solid waste management systems. This system is based on a convolutional neural network and is trained using a custom dataset of images, taken on site from actual moving belts in waste collection facilities. The experimental results indicate that the proposed system can outperform existing algorithms found in the literature in real-world conditions, with 92.43% accuracy.

Finding and recognizing different objects in an image quickly and reliably is an important field of computer vision. Object detection is a challenging problem in this field. Humans have the ability to perform such complex tasks fast and accurately. In contrast, the problem of locating objects via a computer is not so simple. Deep learning algorithms have emerged as powerful methods to detect objects in an image. In this paper, six state-of-the-art object detection algorithms are presented, analysed and compared computationally using four different datasets, two single class and two multiple class datasets. The computational results show that the algorithms achieve higher accuracy on the single class datasets than the multi class datasets.