A Review on Applications of Deep Learning in Agriculture
Agriculture is one of the major industries in the world. As the global population has been continuously increasing, a large increase on food production must be achieved by maintaining at the same time availability and high nutritional quality across the globe, protecting the natural ecosystems by using sustainable farming procedures. However, traditional methods of farming are not enough to handle this huge food demand. This is driving farmers and agro companies to find newer ways to increase production and reduce agriculture waste. As a result, Information and communication technologies (ICT) are emerging as part of the agriculture industry’s technological evolution. Such modern farming technology is called the Internet of Things (IoT) and it does not require the presence of a farmer to control various agricultural processes. IoT based farming system can be used for monitoring the crop field with the help of sensors (light, humidity, temperature, soil moisture, etc.) and automating the farming eco system. The automation in agriculture is the main concern and the emerging area of research across the world. Thus, new automated methods were introduced. These new methods satisfied the food requirements and also provided employment opportunities to billions of people. One of the most prominent technologies for agriculture automation is Deep learning.
The main objective of this paper is to find the various applications of Deep learning in agriculture such as for irrigation, weeding, Pattern recognition, crop disease identification etc. This article performs a survey of 27 research papers that employ different deep learning techniques applied to various agricultural problems, such as disease detection/identification, fruit/plants classification, Automation of irrigation and fruit counting among other domains. The paper reviews the specific employed deep learning models, the source of the data, the performance of each study, the employed hardware and the possibility of real-time application to study eventual integration with autonomous robotic platforms. The current study also compared deep learning with other existing techniques, in terms of performance. Our research findings indicate that deep learning offers better performance and outperforms traditional image processing techniques.