Enhancing agriculture by using artificial intelligence

Rishav Kumar Singh^{* *}Correspondence: Rishav Kumar Singh, Department of Computer Science and Engineering, Raja Balwant Singh Technical Campus Bichpuri, Agra, India, Email:

Author info »

Introduction

The amount of organic matter and total amount of nitrogen in the soil matter plays very important role in soil fertility and its health. These two things affect the physical, biological and chemical properties of the soil. Hence, it is necessary to check if the organic matter and nitrogen is available in satisfactory amount throughout the field. There is one way to take the sample of soil to the laboratory for proximal reflection spectrometry and imaging spectrometry which assess the soil quality and its properties [1]. But the lab tests are not reliable to get an overview of the nutrients’ continuous distribution over a specific area. To tackle this, hyperspectral imaging spectrometry is used to get the data of soil organic matter and nitrogen content over a certain targeted geographical location. It uses hyperspectral sensors’ data from satellites, planes and UAVs [2,3]. Researchers have used the dataset for their projects available from the satellites’ sensors. EO-1 Hyperion satellite’s sensor data is not recent (the data available is till 2014) which makes it not a first choice for recent development in soil properties estimation [4,5]. For new data, GF-5 AHSI data is being used but, satellite based HIS methods will not give accurate results for small-scale application like fruits plant [6-7]. It will be a good choice for covering larger areas. The alternative to satellite based is airplane and UAV based. The airplane based were an ideal choice until UAVs started being used due to its obvious advantages over airplanes like not regulated by air traffic control, cost effective and flexibility

UAVs have been used mostly for classification and mapping functions in precision agriculture. The techniques through which mapping of the field can be done is either passive or active remote sensing. Active remote sensing is (LIDAR) and passive is nothing but optical remote sensing. UAVs can be used for

• Vegetation Indices (VI).
• Plant disease monitoring.
• Crop water stress index monitoring.
• Fertilizers and pesticides spraying.
• Crop phenotyping.
• Weed detection and management.
• Crop yield prediction.

The above Table 1 represents the review and limitations of the research conducted by different authors from the above mentioned Table 1 we can observe that the different technologies used from planting to harvesting while few of them is budget friendly while few are bit costly.

TABLE 1 Research gap/limitation

The another issue is that in order to operate the tools proper training is required to farmers as otherwise it could show the worse effects if not handled carefully.

Materials and Methods

It is evident from the case studies mentioned above that AI has crucial role to play in the applications like Irrigation, yield management, pest and disease control etc. The applications can be categorised as field applications, environmental and agricultural. With the advent of the artificial intelligence, over the years, many new methods have been devised to make the problem solving as easy as possible.

Many of the papers discussed above used AI techniques in one way or another. The subset of the AI is machine learning. The data (statistical o experimental) from the past is fed to the machine to train it and the purpose is to solve any application specific problem.

Some of them are listed below.

• Artificial Neural Networks (ANN)
• Machine learning
• Expert systems
• Reinforcement learning
• Decision support systems
• Deep learning

Various fruit images like apple, guava, kiwi, lemon, cherry etc. were taken in the dataset. Images will be selected and passed through the custom model designed by the researchers. A new method of including pooling layers to the neural networks differs this method from other implementations in the field. The designed model will get 97% accuracy with 19.5 milliseconds. The problem of background information in the crop images interfering the accuracy of the system.

To tackle that, an automatic image segmentation algorithm was developed based on Grab cut algorithm [8]. For deep learning model, mobile net based on CNN was chosen. This algorithm basically removes the background (unnecessary) information (noise) from the crop images and let the model focus on the problem (In this, crop disease identification) [9]. A huge dataset of 38,072 images with 22 diseases from 6 crops were taken and fed to the model. The system was then taken to the smart mobile devices in which the customer can get disease information with accuracy of 84.8%.

All together a high quality micro camera is attached on UAV (drone) which will send its images and spray the color at particular place where the diseased crop or fruit is found the camera on drone is connected with the smart mobile and it will also check the phenotyping of the crop and area where the color is sprayed the farmer could target those area and spray the pesticides or provide water where it is required [10].

Results and Discussion

The widespread implementation of Artificial Intelligence (AI) in agriculture has shown significant results, providing numerous benefits to farmers and the overall agricultural industry. Some notable outcomes include:

Increased crop yields: AI algorithms can accurately predict crop yields based on various factors such as weather conditions, soil quality and historical data. By utilizing these predictions, farmers can make informed decisions regarding planting techniques, resource management and crop rotation, resulting in improved yields.

Enhanced crop health management: AI-powered systems can analyze images of crops taken by drones or sensors to detect early signs of diseases, pests or nutrient deficiencies. This early identification allows farmers to take prompt actions, preventing the spread of infections and reducing crop losses.

Optimized resource allocation: AI algorithms help farmers optimize the use of resources like water, fertilizers and pesticides by providing precise recommendations based on specific crop needs. This reduces waste and minimizes the negative impact on the environment.

Improved efficiency through precision farming: AI enables farmers to implement precision farming techniques by utilizing data analysis and automation. These results in precise planting, targeted irrigation and customized fertilization, leading to higher productivity and cost savings.

Reduced reliance on chemical pesticides: AI-based technologies, such as computer vision and machine learning, can identify and differentiate between crops and weeds, facilitating targeted weed control. Similarly, AI can monitor pest populations and suggest the most effective, environmentally friendly pest control strategies, reducing the reliance on chemical pesticides.

Efficient livestock management: AI systems can monitor and analyze data from sensors attached to livestock, providing valuable insights into their health, behavior and nutritional needs. This data-driven approach allows farmers to optimize feeding schedules, detect diseases early, improve breeding programs and ensure better overall management of their livestock.

Development of autonomous farming equipment: The integration of AI with farming machinery has opened up possibilities for autonomous vehicles and robots. These can perform tasks such as planting, harvesting and data collection with precision and efficiency, reducing labor requirements and increasing productivity.

The Figures 1 and 2 represents the process cycle of agriculture by artificial intelligence here the crop images are used by the sensors and pictographs are used and applied in the dataset, database the neural network and machine learning algorithm worked on that in order to produce a healthy crop by giving idea of adequate amount of fertilizer and pesticides to be used in the crop [11-13].

Figure 1) Representing process cycle of agriculture by AI method

Figure 2) Representing process cycle of plant when AI technique is applied

Figure 3 below compare how the cost used in farming is reduced to 26% in corn, 31% in soybean and 31% in wheat by applying A.I and precision agriculture tools (Table 2). The data represents complete production cycle of crop and where all the cost is reduced [14-16].

Figure 3) Representing reduction in cost of agriculture by AI

TABLE 2 Compare different crops production before and after applying artificial intelligence

As a result of AI in agriculture, farmers can make more accurate decisions, increase productivity, reduce costs, minimize environmental impact and ultimately contribute to food security and sustainability. The ongoing advancements and adoption of AI technologies in agriculture hold significant promise for the future [17].

Conclusion

The integration of AI in agriculture holds tremendous potential for revolutionizing the industry. Different type of UAV like multi-rotor, fixed wing, blimp, flying wing and glider have been designed and used in the fields. Different sensors like multi-spectral image sensors, RGB-D cameras, hyper-spectral imaging sensors, Infrared cameras, LIDAR are mounted on the prototype/commercial drones. They are not limited to Low altitude flying, low-cost, not subject to air traffic control and high operational flexibility. The main challenges during the flight of an UAV are the environmental parameters like solar, wind, soil and noise. These parameters can influence large disparity in data collected which in turn will give less accuracy. The advantage of low altitude flying of UAV brings a problem of crashing which can damage the assembled sensors on it. Thus, low-cost and high-performance sensors are also needed.

References

1. Alencastre-Miranda M, Davidson JR, Johnson RM, et al. Robotics for sugarcane cultivation: Analysis of billet quality using computer vision. IEEE Robot Autom Lett. 2018;3(4):3828-3835.

   [Google Scholar]

2. Latha RS, Sreekanth GR, Suganthe RC, et al. Automatic fruit detection system using multilayer deep convolution neural network. In2021 International Conference on Computer Communication and Informatics (ICCCI). 2021:1-5.

   [Google Scholar]

3. Xiong Y, Liang L, Wang L, et al. Identification of cash crop diseases using automatic image segmentation algorithm and deep learning with expanded dataset. Comput Electron Agric. 2020;177:105712.

   [Google Scholar]

4. Lin P, Chen Y. Detection of strawberry flowers in outdoor field by deep neural network. In2018 IEEE 3rd International Conference on Image, Vision and Computing (ICIVC) 2018:482-486.
5. Syazwani RW, Asraf HM, Amin MM, et al. Automated image identification, detection and fruit counting of top-view pineapple crown using machine learning. Alex Eng J. 2022;61(2):1265-1276.

   [Google Scholar]

6. Latha RS, Sreekanth GR, Suganthe RC, et al. Automatic fruit detection system using multilayer deep convolution neural network. In2021 International Conference on Computer Communication and Informatics (ICCCI) 2021:1-5.

   [Google Scholar]

7. Zhang Y, Phillips P, Wang S, et al. Fruit classification by biogeography‐based optimization and feedforward neural network. Expert Syst. 2016;33(3):239-253.

   [Google Scholar]

8. Liu LW, Ma X, Wang YM, et al. Using artificial intelligence algorithms to predict rice (Oryza sativa L.) growth rate for precision agriculture. Comput Electron Agric. 2021;187:106286.

   [Google Scholar]

9. Reyes-Yanes A, Martinez P, Ahmad R. Real-time growth rate and fresh weight estimation for little gem romaine lettuce in aquaponic grow beds. Comput Electron Agric. 2020;179:105827.

   [Google Scholar]

10. Wikipedia contributors. Agricultural machinery. Wikipedia, The Free Encyclopedia. 2021. Web. 2021.
11. Faiçal BS, Costa FG, Pessin G, et al. The use of unmanned aerial vehicles and wireless sensor networks for spraying pesticides. J Syst Arch. 2014;60(4):393-404.

    [Google Scholar]

12. Faical BS, Freitas H, Gomes PH, et al. An adaptive approach for UAV-based pesticide spraying in dynamic environments. Comput Electron Agric. 2017;138:210-223.

    [Google Scholar]

13. Zhu W, Sun Z, Huang Y, et al. Optimization of multi-source UAV RS agro-monitoring schemes designed for field-scale crop phenotyping. Precis Agric. 2021;22(6):1768-1802.

    [Google Scholar]

14. Comba L, Biglia A, Ricauda Aimonino D, et al. Leaf area index evaluation in vineyards using 3D point clouds from UAV imagery. Precis Agric. 2020;21:881-896.

    [Google Scholar]

15. Saha AK, Saha J, Ray R, et al. IOT-based drone for improvement of crop quality in agricultural field. In2018 IEEE 8th Annual Computing and Communication Workshop and Conference (CCWC). 2018:612-615.

    [Google Scholar]

16. Boursianis AD, Papadopoulou MS, Diamantoulakis P, et al. Internet of things (IoT) and agricultural Unmanned Aerial Vehicles (UAVs) in smart farming: A comprehensive review. Internet of Things. 2022;18:100187.

    [Google Scholar]

17. Feng L, Chen S, Zhang C, et al. A comprehensive review on recent applications of unmanned aerial vehicle remote sensing with various sensors for high-throughput plant phenotyping. Comput Electron Agric. 2021;182:106033.

    [Google Scholar]