Application and prospect of black technology-artificial intelligence identification technology in tea plant diseases and insect pests identification

Artificial intelligence recognition technology has been widely used in industry, agriculture and other fields because of its fast recognition speed, good stability and high accuracy. In recent years, scholars began to apply artificial intelligence identification technology to the identification of tea plant diseases and insect pests, in order to improve the identification efficiency and save labor.

First, the development of artificial intelligence identification technology

The exploration of artificial intelligence identification technology began with the study of biological vision in the 1950s. Image acquisition equipment is generally used to automatically receive the target image and process and analyze the image, which has the characteristics of high speed, good stability and high precision, and has the development potential of replacing human eyes for identification.

Technology-application and prospect of artificial intelligence identification technology "img _ height =" 411"img _ width =" 985 "data-src ="//imgq8.q578.com/ef/0719/7673af3. After entering the 2 1 century, traditional machine learning methods and deep learning have been widely used in the research of artificial intelligence to identify agricultural pests and diseases. The early research is based on static sample images, and the recognition effect needs to be improved in complex field environment. Deep learning has certain advantages in dealing with massive data, which can automatically extract object features from large-scale data and use classifiers for classification and recognition. Compared with traditional machine learning, deep learning has obvious improvement in recognition accuracy and efficiency, and has obvious advantages in improving recognition accuracy and reducing R&D labor input.

Second, the research status of artificial intelligence to identify tea plant diseases and insect pests

1. Research progress of artificial intelligence to identify tea plant diseases and insect pests

According to statistics, there are more than 900 recorded tea plant diseases and insect pests in China. In the past, the identification of these tea plant diseases and insect pests mainly relied on plant protection experts and plant protection workers, through the morphological characteristics, occurrence characteristics and occurrence time of pests. Traditional manual identification is difficult to meet the production demand, which brings difficulties to accurate prevention and control. In contrast, artificial intelligence recognition is obviously more accurate and takes less time. Therefore, the application of artificial intelligence in the identification of tea plant diseases and insect pests has great potential and demand.

With the development of artificial intelligence identification technology in agricultural pest identification system, the research on tea tree pest identification has made some progress. In 2008, on the basis of expert experience, Qin developed 1 set of intelligent WEB management system for tea garden pests, which includes three main links: tea garden pest identification, pest prediction and tea garden pest control decision-making, and uses morphological identification, map identification and retrieval identification to identify pests, which is a representative study of introducing artificial intelligence technology into tea garden pest control in early China. In the field of pattern recognition, the algorithm has an important influence on the recognition speed and the accuracy of the results. Wu et al. used BP, CART and other three algorithms to construct the three-dimensional spatial structure knowledge base of five kinds of tea geometrid pests, and the classification recognition rate of this pest was between 80.00% and 86.67%.

In recent years, convolutional neural network technology has been widely used in the field of image artificial intelligence recognition. The model established by image saliency analysis and convolutional neural network is used to identify common pests in tea garden, which has achieved good recognition effect and improved the recognition ability of different tea tree disease images. The rapid popularization of mobile intelligent devices also provides a feasible direction for the development of pest identification.

At present, the Tea Research Institute of Chinese Academy of Agricultural Sciences has cooperated with Hangzhou Ruikun Technology Co., Ltd. to develop an intelligent identification system based on mobile terminals, which can identify about 80 kinds of common pests and natural enemies in tea gardens. The operation is simple, the identification speed is fast and the accuracy is high, which provides a reliable way for the diagnosis of tea plant diseases and insect pests.

Technology-application and prospect of artificial intelligence identification technology "img _ height =" 860 "img _ width =" 860 "data-src ="//imgq8.q578.com/ef/0719/4854e62adc5438+0.jpg "src. Technology-application and prospect of artificial intelligence identification technology "img _ height =" 710 "img _ width =" 360 "data-src ="//imgq8.q578.com/ef/0719/d91c5. 2. The problems of artificial intelligence in identifying tea plant diseases and insect pests.

In the past few decades, artificial intelligence identification technology has developed rapidly. The application of deep learning in the field of pest identification and the optimization of various algorithms have greatly improved the efficiency and accuracy of pest identification, but there are still many problems in the development of artificial intelligence in the study of tea tree pest identification.

On the one hand, most of the research is still in the stage of laboratory research, which can not meet the requirements of practical application. The main reason is that most of the current research is conducted indoors, which can effectively exclude the influence of other external interference factors. However, in practical application, the environment of tea garden is complex, and illumination and weather will have a certain impact on the collection of pictures, and pests and diseases will be blocked by tea leaves and buds, which will have a certain impact on the identification effect; Second, laboratory research mainly focuses on static pests or disease specimens, but in practical application, it is difficult to identify dynamic tea garden pests, and the accuracy of identification needs to be improved; Thirdly, the pictures collected in pest identification research are mainly in the obvious stage of pest occurrence, but in the early stage of pest occurrence, it plays an important role in taking correct control measures, which requires the identification ability of pest identification system to cover the complete occurrence process of pests.

On the other hand, the development of recognition software should be lightweight, simple, convenient and easy to operate, so as to facilitate the integration of various technical means.

Technology-application and prospect of artificial intelligence identification technology "img _ height ="1026 "img _ width ="1246 "data-src ="//imgq8.q578.com/ef/0719/68d655. Third, the application prospect of artificial intelligence identification technology in the identification of tea plant diseases and insect pests

Although there are still some problems in the application of artificial intelligence identification technology in tea plant diseases and insect pests identification, at present, artificial intelligence identification technology has achieved corresponding results in the design and implementation of tea plant diseases and insect pests identification scheme. In the future, based on this achievement, we can develop in the direction of monitoring, early warning and precise prevention and control of pests and diseases, thus promoting the construction of digital tea gardens.

In the aspect of monitoring and early warning of tea garden pests, with the improvement of effective algorithms, the accuracy of pest identification and the ability to classify the degree of pest damage will be greatly improved. Through intelligent identification of tea plant diseases and insect pests, classification of damage degree and other methods, the single intelligent identification of tea plant diseases and insect pests is gradually transferred to intelligent monitoring and early warning of various tea plant diseases and insect pests, and the advantages of artificial intelligence are fully exerted to realize real-time, dynamic and comprehensive monitoring and early warning of tea plant diseases and insect pests, continuously optimize the monitoring and early warning level of tea plant diseases and insect pests, and provide reliable data for monitoring and early warning of tea plant diseases and insect pests.

In the precise prevention and control of tea garden pests, the database of tea garden pests and natural enemies was established through long-term and multi-point intelligent monitoring data of tea garden and local geographical location. When pests and diseases break out, timely push the occurrence of tea garden pests and diseases according to the local geographical location, climate, natural enemies and monitoring and early warning information, provide accurate pest control measures for tea farmers, avoid the indiscriminate use of pesticides by tea farmers, and promote the popularization of green prevention and control in tea gardens.

The identification system of tea plant diseases and insect pests is an important part of digital tea garden. In the future, the intelligent identification system of tea garden is not only limited to pests and diseases, but also can be extended to tea cultivation, tea garden management and other aspects, from single identification of tea pests and diseases to identification and monitoring of tea growth, cultivation and other aspects, so as to realize the integration of multiple functions into one system and improve the digital management level of tea garden.

This paper is taken from China Tea (No.6, 2022), P 1-6, Application and prospect of artificial intelligence identification of tea plant diseases and insect pests. Authors: Yang Fengshui, Wang Zhibo, Wang Weitong, Zhang Xinxin, Sun Liang, Xiao Qiang.