A Methodology for Early Detection of Plant Diseases Using Real Time Object Detection Algorithm

• Published in: Smart and Sustainable Agriculture Vol. 1470; pp. 122 - 139
• Main Authors: Walia, Ranjeet, Sharma, Shivam, Shrivastava, Swapnil
• Format: Book Chapter
• Language: English
• Published: Switzerland Springer International Publishing AG 2021; Springer International Publishing
• Series: Communications in Computer and Information Science
• Subjects: Deep learning; Plant disease detection; Yolo algorithm
• ISBN: 9783030882587; 3030882586
• ISSN: 1865-0929; 1865-0937
• DOI: 10.1007/978-3-030-88259-4_9

The convergence of Big Data Analytics, Deep Learning, Internet of Things (IoT) and Cloud Computing has catalyzed the emergence of novel applications in various domains. The growing need for Sustainable Agriculture and advancement in these emerging technologies has enabled applications like Precision Agriculture and Smart Farming. There are several such initiatives where agricultural data is being collected continuously by means of devices (e.g., sensors, cameras) and stored in a suitable Big Data Analytics platform. There is a need to generate actionable insights in real time for these applications using various analytical techniques including Deep Learning. Various techniques have been used for plant disease detection, pest detection, plant growth assessment, yield prediction and so on. As per the Food and Agriculture Organization of the United Nations, there is a growth in frequency of occurrences and complexity of plant diseases. In this work we have attempted to build a contemporary model with available resources for early detection of plant diseases. The freely available plant images dataset was collected from various sources followed by required pre-processing to prepare a large, assorted and robust plant disease dataset. The real time object detection algorithm is applied for efficient and timely plant disease detection. The detailed methodology for creation of material, building the detection model and inferring the result is discussed in this paper. The early detection of plant diseases would help in quick diagnosis and provisioning of remedial measures to the farmers for enabling timely preventive or curative actions.