The United Nations Food and Agriculture Organization (FAO) reported that the burgeoning global population will be around two billion by 2050, while only 4% additional land will come under cultivation by then. It’s an uphill task for the farming community to feed the ever increasing world population amid rising agricultural debts, unpredictable weather patterns and biotic stress.
Insect pests and diseases are one of the major reasons for decreasing farm productivity causing 20 to 40 percent global crop loss every year.
In the absence of knowledge and expertise, farmers are over-dependent on pesticide dealers for support on pest identification and their management, which results in excessive and injudicious use of pesticides for controlling the pests. The major concern of farmers for decision making in pest management is “pest identification and timely availability of correct pest management information”. To detect plant pests at an early stage and save undesirable consumption of pesticides, advanced technical solutions are needed in agriculture which will result in saving crop worth crores of rupees or in non-application of intervention saving the cost of intervention involved and thus saving the environment. The core of the pest management framework is the decision-making process. Decision-making in pest management is a dynamic and complex process that requires much more knowledge and support than conventional agriculture.
Pest identification and availability of correct management information are the vital aspects of process of decision-making in pest management. Eye/physical observation methods have been used in recent years, but they are not efficient. The future of farming depends largely on adoption of cognitive solutions. Hence Artificial Intelligence (AI) plays a major role which can greatly help in efficient and successful crop pest management.
Artificial Intelligence (AI) and its role in agriculture
Artificial Intelligence (AI) is a branch of computer science which deals with the simulation of human intelligence processes by computer systems. AI is becoming pervasive very rapidly due to its robust applicability to solve several problems which cannot be done by traditional computing and human efforts. AI possesses the capability to learn from data and thus identify the patterns in the data more efficiently than humans, enabling researchers to gain more insight out of their data. AI is in its nascent stage and will be playing a huge role in futuristic Agriculture scenario of World by the following measures,
Real time Crop and soil monitoring.
Crop yield prediction and price forecast.
Pest identification and timely spraying.
Making resource allocation wiser.
Improving food & environmental sustainability
Analzing market demand and managing risk
Protecting, feeding and harvesting the crops.
Role of Artificial Intelligence in pest management
Plant protection is an extremely important aspect of agriculture to boost crop production and thereby food security. The plant protection measures are to be taken on a community basis so as to ensure effective management of pests and hence Artificial Intelligence (AI) techniqueshas been recently introducedfor precision control of plant insect pests. There are different ways of AI in pest management,which are described as follows.
Easy method for scouting fields: AI can help the scouts in providing accurate descriptions of pest and their exact location in fields.
Addressing challenges in diagnosis of pest: Proper identification of specific pest in the field is important for its successful management. Another important aspect of pest management is regular pest monitoring, which helps to determine the level of incidence and timing to initiate pest management intervention.
Predicting pest problems early: Application of AI techniques can help to automate and speed up the process of providing timely and correct decision-support to the farmers on important aspects of pest management such as pest identification, pest monitoring and selection of appropriate pest management strategy
Large-scale pest monitoring and surveillance:Drones which work on principles of artificial intelligenceare used forpest monitoring, surveillance.
Pest management: Spraying of pesticides by AI based drones to control pest efficiently over a larger area by ensuring complete coverage of crop.
AI techniques for crop protection
Machine learning is concerned with algorithms that can learn on their own from a given collection of input data in order to achieve a specific goal. Its high-performance computer opens up new possibilities in agriculture. In the agricultural domain, machine learning and statistical pattern recognition have sparked a lot of attention because they promise to improve the sensitivity of disease detection and diagnosis. Machine learning-enabled solutions give farmers a wealth of recommendations and insights to aid in decision-making and action.Example: Classification of diseased or non-dispersed leaves, fruit, plants, etc.
2.Artificial Neural Network (ANN)
ANNs are one of the more reliable ways of identifying plant diseases among the several approaches employed (ANNs). In order to improve feature extraction, neural networks are integrated with various image pre-processing algorithms. ANN is based on biological neurons in the human nervous system. ANN, on the other hand, can infer meaning from complex data and uncover patterns that are too tough for people or traditional computers to detect. Other advantages of ANNs include adaptive learning, self-organization, real-time operations, and so on.
3.Image Processing Techniques
For effective detection and classification of the plant, image processing techniques were widely and successfully applied. A two-dimensional taxonomy is used to categorise the data. Object recognition, data reduction/feature extraction, pre-processing, segmentation, optimization, and image interpretation are all part of one dimension. In a distinct dimension, inputs are received and tasks are accomplished at various levels, such as pixel level, object set level, and so on. For boosting the efficiency of illness diagnosis, several pre-processing techniques such as picture clipping, image smoothing, and image enhancement are used. Image segmentation can be accomplished using a variety of techniques, including the Otsu method, k-means clustering, and transforming RGB images to HIS models. Fourier filtering, edge detection, and other image pre-processing techniques were used.
Example: Image based disease and weed identification.
4.Support Vector Machine (SVM)
A supervised learning system called a support vector machine is used to solve classification and regression problems. The hyperplane is used to distinguish the classes in SVM. In N-dimensional space, a hyperplane is comparable to a line in two-dimensional space. This hyperplane is a line in two-dimensional space that divides a plane into two halves, with each class on either side. The SVM method uses labelled training data to find the optimum hyperplane for categorising fresh samples. As a result, the hyperplane is found by SVM to categorise the data points individually. The Support Vector Machine (SVM) has also been found to be very promising for accurately classifying leaf diseases.
Internet of Things (IoT): The internet of things, or IoT, is a system of interrelated computing devices, mechanical and digital machines, objects, animals or people that are provided with unique identifiers and the ability to transfer data over a network without requiring human-to-human or human-to-computer interaction. Sensors and robotics are the part of IoT. Example: Robotics (Drones) helps to take the view or infestation survey of the field within a short span of time without manual power.
Conclusion and future challenges
The main prospect of using artificial intelligence is pest monitoring, identification and timely recommendation of plant protection measures. It is the latest way farmers can adopt new technology to meet the global food demands by managing insect pests through artificial intelligence techniques and hence contribute to the increase in food security. Many mobile apps based on artificial intelligence are been developed by different ICAR- research institutes for different crops to identify and manage the insect pest of crop efficiently.Although use of AI is promising, there are challenges when it comes to plant protection. The development of innovative AI algorithms and non-availability or limited availability of data for data learning are two major challenges in the process of developing AI based plant protection tools and techniques. Pest prediction is still complex and elusive. Process of plant protection in agriculture is slowly becoming digital with AI showing promising potential.
Niranjan Singh1, M. K. Khokhar1,Licon Kumar Acharya1 T.K Mondal2 and Shbana Begam2
1 ICAR-National Centre for Integrated Pest Management, New Delhi
2ICAR-National Institute for Plant Biotechnology, New Delhi