Agriculture is the main source of livelihood in most of the developing countries of the world. Studies indicate that around 60 percent of the world’s population is dependent on it for survival. Increased agricultural production can also significantly boost the economic development of a country. However, achieving desired levels of productivity is a challenge as it is influenced by various factors like climate, pests, diseases amongst many others.
Disease is one of the significant factors that can adversely affect the yield of a crop. If left unchecked, it will eventually lead to severe financial losses to farmers. Maize, more popularly known as corn is among the most consumed food crops in the world. It is the world’s most produced grain in terms of volume with around 1100 million metric tons produced last year. Countries like India, China and USA are amongst the top cultivators of maize in the world. Like every other crop, it is also affected by diseases like leaf blight, common rust, stalk rot, holcus spot, etc. which impedes plant growth, resulting in decreased production. The humongous volume of production dictates the need for early detection of diseases in these plants and can prove helpful as appropriate diagnosis at the initial stage can save the yield.
Modern technologies like the pattern recognition feature of machine/deep learning algorithms can be used to study images of leaves and create classification models. This trained model can then be used to detect diseases in future at an early stage. It provides ample time for farmers to address the issue and treat the affected plants. This saves the crop from destruction and will result in increased production, thereby helping the agricultural society. It also eliminates the need for naked eye observation of experts which can be an expensive and time-consuming exercise (Amoda, Jadhav and Naikwadi, 2014).
Save your time!
We can take care of your essay
- Proper editing and formatting
- Free revision, title page, and bibliography
- Flexible prices and money-back guarantee
Justification
Past researches help us understand the complexity and challenges faced by researchers in this area. Initial studies in this area revolved around the usage of image processing techniques like histogram and edge detection. Amoda et al. (2014) made use of these methodologies to classify leaf diseases and achieved acceptable results using MATLAB.
Jagan, Balasubramanian and Palanivel (2016) employed machine learning algorithms like k-Nearest Neighbor (kNN) and Support Vector Machine (SVM) to tackle the problem and made significant discoveries supporting the usage of these methodologies. Aravind et al. (2018) employed similar algorithms along with Bag of Features to detect leaf diseases in maize plant. The results obtained were about 80 percent accurate leaving scope for improvement.
Authors (Li, Jia and Xu, 2018) have made use of unsupervised learning and deep learning techniques like convolutional generative adversarial networks for image-based plant disease detection and have achieved accuracy rates of around 90 percent which are significantly higher than previous researches in this area.
As supported by earlier studies, machine learning algorithms are highly effective and can be used to solve problems in this domain. In this research, one shall try to make use of deep learning algorithms like transfer learning, multilayer perceptron (MLP), convolutional neural networks (CNN) and its variants to identify and classify diseases in maize plant. These algorithms, unlike the ones that were previously used by researchers are specifically meant for analyzing image-based data and may help us in achieving better results. Tools like R/Python and SPSS shall be used to implement the same. Publicly available datasets will be used in the research thereby preventing any violation of ethics.
References
- Amoda, N., Jadhav, B. and Naikwadi, S. (2014) ‘DETECTION AND CLASSIFICATION OF PLANT DISEASES BY IMAGE PROCESSING’, International Journal of Innovative Science, 1(2).
- Aravind, K. R., Raja, P., Mukesh, K. V., Aniirudh, R., Ashiwin, R. and Szczepanski, Cezary (2018) ‘Disease classification in maize crop using bag of features and multiclass support vector machine’, in Proceedings of the 2nd International Conference on Inventive Systems and Control, ICISC 2018. doi: 10.1109/ICISC.2018.8398993.
- Jagan, K., Balasubramanian, M. and Palanivel, S. (2016) ‘Detection and Recognition of Diseases from Paddy Plant Leaf Images’, International Journal of Computer Applications. doi: 10.5120/ijca2016910505.
- Li, J., Jia, J. and Xu, D. (2018) ‘Unsupervised representation learning of image-based plant disease with deep convolutional generative adversarial networks’, in Chinese Control Conference, CCC. doi: 10.23919/ChiCC.2018.8482813.
Stuck on your essay?
