Ontologies are being used in systems to represent knowledge in various AI technologies and model different CPS ecosystems where smart farming has transformed the agricultural field which has helped create the co-op necessary for cooperative success. The system has helped farmers in making decisions during extreme weather events that are frequent and critical for agricultural sectors. Smart Farming has revolutionized agriculture, which in turn has helped increase the quantity and quality of food and raw materials. Establishing a legal and contractual entity in the early stages of a cooperative is necessary for the success of the cooperative because abuse of the cooperative's services and common resources by its members may not achieve its goals and ultimately has led to disintegration. Autonomous mobile robots are also tools in precision agriculture to perform various tasks. Control robots can adapt and learn, which is critical for agriculture, which is a powerful strategy.
Most autonomous robots have sensors that input data, which is then processed by a control unit. The robot control system can be based on fuzzy logic. Robots can be used to inspect and handle plants thanks to built-in touch and optical sensing systems. Some other widely used robotic applications are weed picking and robotic weed control, which are based on machine vision and include accurate chemical applications. This seems very practical as manual weed control is a very tedious and inefficient task that adds to human labour. In addition to this, robots have been used for phenotypic plant health assessment. While different robots use different navigation systems, they are usually guided by GPS and a human-controlled laptop as they move between rows of plants. Similarly, there has been progress in using robots to harvest crops such as apples, grapes and others. Well-established AI techniques, when applied to data collected from field sensors, can help develop more efficient data-driven intelligent farming systems. They used the described neural networks to describe plant stress severity based on visual models. The AI seeding app is developed by Microsoft. The system makes recommendations such as the optimum timing of seed sowing, soil preparation for cultivation, etc. Lee et al. developed a tool that helped identify the pest hazards of fruit trees. Tools like AutoTrac use AI techniques to evenly plant crops to reduce overlap and excessive plant spacing. Blue River Technology has used computer vision techniques to identify individual plants and identify anomalies. Ca et al., detail a method for classifying dense grasslands from aerial photographs using a cascaded convolutional neural network encoder-decoder.
Another AI approach showing promising results is the use of support machines for sorting and fuzzy logic to automatically sort agricultural products without human intervention. Diagnosis of diseases in plants Plants are very susceptible to disease as they are ecosystems, so disease prevention and control is very important. Current crop conditions and severity of infection affect the rate of spread of the disease. The key to preventing loss of crops and agricultural products is the detection of plant diseases. Colorful spots or stripes that can appear on the leaves, stems, and seeds of a plant are several signs that the plant is sick. Therefore, early diagnosis remains elusive in many parts of the world. Advances in computer vision through deep learning have paved the way for disease diagnosis using smartphones. The tedious task of observing huge crop fields and detecting disease symptoms at an early stage is very tedious, so automated systems are useful. Therefore, the aim was to use a software program based on image processing to detect and classify plant root diseases automatically.
Patil and Kumar (2011) aimed to provide a variety of advanced methods for plant disease/trait research using image processing to increase productivity and reduce costs associated with bringing in specialists in plant disease diagnosis. Identification of diseased leaves, leaves, fruits, quantification of diseased area, morphology of affected area, color of affected area, determination of size and shape of fruits, etc. Image processing is useful. A manual evaluation script that changes the quality limiting step to imaging can be extended beyond its feasibility study by automating an image analysis experiment. Several methods and approaches can be used to classify and diagnose diseases using computer vision. Deep convolutional neural networks were used, achieving 99.53% success in diagnosis with the corresponding plant. Neural networks have also been used to detect diseases in crops such as rice. K-Means Algorithm, Principal Component Analysis (PCA), Coefficient of Variation (CV), Support Vector Machines (SVM) are also some of the other options and in context some model method is more efficient. In the illustrative study, K-means clustering classification into two groups: healthy and infected, followed by assistive machines (SVMs), produced better results than ANNs. All the visible characteristics of an organism resulting from the interaction of its species (complete genetic inheritance) with the environment can be described as phenotyping.
Traits may include behavioral, chemical, color, shape, and size. Collection, analysis and application of plant statistics remain inadequate. Furthermore, the phenotype reflects an enormous number of processes, functions and structures that change during the course of growth and development. A prerequisite for breeding, introduction of varieties, genomic and phenomics studies is a thorough evaluation of crop varieties. Increasing the yield is the main objective and problem of plant breeding. Dee and French (2015) sought to propose an automated system based on computer vision that can perform feature identification and measurement without human intervention, in which we can obtain higher output and higher accuracy in less time and even at lower cost. Most cost traditional ones. function. According to Coppens et al. (2017) Robotic imaging systems can significantly increase the throughput volume of characterization, thus overcoming the perceived drawback of gene design. Thus, the efficiency of new phenotyping and genotyping methods should be evaluated considering the relative genetic gains that can be obtained through the introduction of new methods, whereas the cost-benefit should be evaluated in relation to cost and cost of additional genetic benefits.