Technical Field

The invention relates to an artificial intelligence-based predictive decision support system in disease, pest and weed control.

In particular, the invention relates to an artificial intelligence-based predictive decision support system in the control of diseases, pests and weeds, which is used to economically protect plants from the damage of diseases, pests and weeds that limit plant production, and thus to increase agricultural production and improve their quality.

Background of the Invention

Meeting the food needs of people is one of the most important issues that countries focus on today. Despite all efforts, the world population is increasing and the surface area of the world is not increasing. In addition, agricultural areas are gradually decreasing due to erosion, opening of new settlements, industrial facilities and roads. Since it is not possible to increase the amount of land, it becomes necessary to increase the amount of product obtained from the unit area by using modern techniques and inputs. Among the ways to increase the yield are irrigation, proper tillage, fertilization, improvement, appropriate harvest, establishment of producer associations, mechanization as well as the application of modern plant protection methods.

There is a decrease in yield due to diseases, pests and weeds in agricultural areas in our country. This leads to serious losses both in the income of our producers and in the gross national product of our country. Therefore, plant protection studies must be applied correctly, at the right time and with the right diagnosis. Countries are faced with purchasing agro-industrial products to meet their food needs. Herbal products and various plant-derived substances are displaced from country to country, so that plants and herbal products are dispersed all over the world in a short time by crossing national borders.

Today, there is a rapid and large-scale exchange of plants in the world. As a result of this, plant and plant product parts are dispersed to far distances in a short time. Along with these, very dangerous diseases, pests and weeds are distributed. If no precautions are taken, clean countries and regions are infected with harmful factors in a short time. These factors, which took a long time to spread before, are now contagious and dispersed in a short time.

Since any non-economic application has no place in modern plant protection, the aim of agricultural struggle today is to increase the product and quality within the limits of economy. As it is known, economy includes the most fundamental values of our age.

The concept of economy in modern plant protection works not only to protect the environment and health, but also to prevent new problems, such as new pests, diseases or weed species from becoming dominant, with conscious and controlled practices.

Diseases, pests and weeds in agricultural areas cause significant losses in yield and reduce the profitability of the market by reducing product quality. The producer sometimes mixes up well-known diseases, pests and weeds, and agricultural engineers and consultants working in the field have difficulty in diagnosing many diseases and pests precisely.

Producers unconsciously spray diseases, pests and weeds, and as a result, they cannot get the desired product. Plant protection products applied unconsciously harm the soil, water resources, natural flora, animals and plants. Producers often apply pesticides unnecessarily by imitating the pesticide application in the neighboring parcel without asking the agricultural engineers. In the Chinese patent application CN107942302 in the literature, regarding the subject, “The invention discloses a smart radar sea clutter forecasting system and method based on an invasive weed optimization algorithm. The system is composed of a radar, a database, and a host computer which are connected in sequence. The radar irradiates a detected sea area and stores radar sea clutter data in the database. The host computer includes a data preprocessing module, a robust forecasting model modeling module, an intelligent optimization module, a sea clutter prediction module, a discrimination model updating module, and a result display module. According to the invention, as for the chaotic characteristics of the radar sea clutter, the radar sea clutter data are reconstructed, the reconstructed data are subjected to nonlinear fitting, and the invasive weed optimization method is introduced, so as to establish an intelligent forecasting model of the radar sea clutter. The invention provides the intelligent radar sea clutter forecasting system and method based on the intrusive weed optimization algorithm which avoid artificial factors and have high intelligence.” statements are included.

The application mentioned is about a radar system on invasive weed in the seas. Radar creates an intelligent forecasting model based on data on the chaotic properties of marine clutter.

In the patent application numbered CN111387169 in the literature, "The invention describes a targeted precision sorting module with a controlled spray range, a multispray path based on artificial intelligence technology and is related to the field of agricultural machinery. Multi-spray path based on artificial intelligence technology, targeted precision sorting module with controlled spray interval, especially applied to intelligent field sorting of large number of crops in agriculture. The module is used for multiple types of transport platforms and quick snap assembly can be provided. The device is suitable for field weeding of a large number of field crops at different growth cycles. According to the device, a light convolutional neural network is adopted to recognize weed information, thus providing the basis for agent type selection, target spray parameters and flows, automatic optimization of pressure, spray head angles, and deposition zone planning. A multi-channel target spray unit can spray different types of herbicides on different types of weeds, a synchronous belt slide table moves the spray heads transversely to accurately target spray on the weeds, and a spray interval can be adjusted according to the following. The size of weed groups. Through the targeted spraying system, pesticide use can be reduced, pollution to an ecological environment can be reduced, and working efficiency can be effectively improved.” statements are included.

In the aforementioned patent application, on the other hand, the device works only according to the growth differences, not according to the species characteristics designed for weeding a wide variety of field crops in the field at different growth cycles. It is designed as an agricultural machine so that it can spray different types of herbicides on different types of weeds. It's just about weed.

Again, in the patent application numbered CN109676624 in the literature, "The invention aims to provide a field target detection robot platform based on deep learning. The field target detection robot platform consists of a vehicle body and an artificial intelligence system group, where the vehicle body consists of a frame platform, a walking and guidance system, a double-wing vision mechanism and an intelligent control system. Power is provided by a hub motor in the drive and steering system. A stepper motor is used to control the forward direction of the platform to perform a steering function. Based on the aforementioned, a camera is used in the double-bladed vision mechanism to collect field plant and weed information in the artificial intelligence system and to transmit this information to the computer to detect and classify the detected plants. The intelligent control system in the vehicle body is used to control the walking and steering system and a steering system. Meanwhile, the vehicle body is subject to speed regulation and progressive direction change according to the instructions of the artificial intelligence system.” statements are included.

In the mentioned application, it is related to an area target detection robot and is used for collecting field plant and weed information and transmitting this information to the computer to detect and classify the detected plants. It just takes a camera image from the plant and sends it to a computer in the center.

Due to the disadvantages mentioned above, there was a need to introduce a new artificial intelligence-based predictive decision support system in disease, pest and weed control. Disclosure of the Invention

Based on this position of the technique, the aim of the invention is to introduce a new artificial intelligence-based predictive decision support system in disease, pest and weed control that eliminates the existing disadvantages.

Another aim of the invention is to present a system that limits plant production, protects plants economically from the damage of diseases, pests and weeds, and thus increases the production quality by increasing agricultural production.

Another aim of the invention is to present an application that agricultural engineers and farmers can also benefit from and can use, as well as automatic regional mapping of diseases, pests and weeds, which are the most challenging and economically challenging for our farmers.

Another aim of the invention is to introduce a dynamic system that performs regional mapping and forecasting as it is used by the end user (agricultural engineers and farmers) and data expands.

Explanation of Figures

Figure - 1 A schematic view of the artificial intelligence-based predictive decision support system in disease, pest and weed control, which is the subject of the invention

Reference Numbers

A- Artificial Intelligence-Based Predictive Decision Support System In Disease, Pest And Weed Control

1 . Disease, Pest and Weed Formations

2. Diseased Fruit/Veg etable

3. Application

4. Raw Data

5. Artificial Intelligence Engine

6. Learned Database 7. Main Databases

8. Classified Data

9. Expert Pool

10. Feedback Unit

Detailed Description of the Invention

In this detailed explanation, the innovation that is the subject of the invention is only explained with examples that will not have any limiting effect for a better understanding of the subject.

The invention is an artificial intelligence-based predictive decision support system (A) in the control of diseases, pests and weeds, which is used to economically protect plants from the damage of diseases, pests and weeds that limit plant production, and thus to increase agricultural production and improve their quality, characterized in that; comprises the process steps that, detection of images by photographing or analyzing disease, pest and weed formations (1 ) and/or diseased fruits/vegetables (2), creating raw data (4) by uploading the detected image to the application (3) with image processing, sending the raw data (4) of the incoming photos to a central server where the artificial intelligence engine (5) is running, analyzing disease, pest and weed formations (1 ) available in main databases (7) by deep learning trained with learned database (6) and recording it as classified data (8) with the data processing technique in the software, with the determinations of the subject experts, obtaining classified data (8), which is learned artificial intelligence model data, by analyzing the images uploaded in the main databases (7) of the artificial intelligence engine (5), which includes the artificial intelligence-based system, in the algorithmic analysis module and determining the differences and recording the obtained classified data (8) which is the learned artificial intelligence model data in the learned database (6).

In Figure-1 , a schematic view of the artificial intelligence-based predictive decision support system (A) in disease, pest and weed control, which is the subject of the invention, is illustrated. Artificial intelligence-based predictive decision support system (A) which is the subject of the invention, consists of the main elements of, disease, pest and weed formations (1), diseased fruit/vegetable (2), application (3), raw data (4), artificial intelligence engine (5), learned database (6), main databases (7) and classified data (8).

In the application of the invention, the selected disease, pest and weed formations (1 ) and the diseased fruit/vegetable (2) are photographed or analyzed and the image is uploaded to the application (3) with image processing and raw data (4) is formed. In addition, photos of diseases, pests and weeds in the main databases (7) of the Ministry of Agriculture and Forestry and Universities are also used.

The raw data (4) of the incoming photographs are sent to a central server where the artificial intelligence engine (5) is running, and disease, pest and weed formations (1 ) available in the main databases (7) are analyzed by deep learning trained with the learned database (6) and It is recorded as classified data (8) with the data processing technique in the software with the determinations of the subject experts.

With the artificial intelligence engine (5), new images are analyzed within the system and controlled by experts. Artificial intelligence engine (5), which includes an artificial intelligence-based system; by analyzing the images uploaded in the main databases (7) and detecting differences in the algorithmic analysis module, classified data (8), which is the learned artificial intelligence model data, is obtained and the classified data (8), which is the learned artificial intelligence model data, is recorded in the learned database (6).

The main databases (7) work in relational and no-sql (unrelated) architecture and are prepared for Big Data retention. It analyzes by working with computer-vision algorithms and artificial intelligence algorithms included in the image processing artificial intelligence algorithm module in the mentioned artificial intelligence engine (5). Evolving neural networks and CNN Algorithm methods are used to find the best result in the development process of artificial intelligence.

The artificial intelligence-based predictive decision support system (A) in disease, pest and weed control, which is the subject of the invention, is trained with species and breed differences that can detect both disease, pest and weed formations (1) and diseased fruits/vegetables (2) in agricultural production. It will analyze by working with computer-vision algorithms and artificial intelligence engine (5) based artificial intelligence algorithms by using learned databases (6) with deep learning software system.