Disease and pest recognition system based on big data and deep learning

Abstract

The invention discloses a disease and pest recognition system based on big data and deep learning. The system comprises a data collection module for collecting data by adopting a crawler task scheduling platform based on a Hadoop framework, a data storage and processing module for effectively storing and processing the data by using a MySQL database in combination with an HDFS, and a data application module for improving a classic YOLO V4 algorithm for a disease and pest classification task. By using the task scheduling platform, a large amount of data on the Internet can be conveniently obtained, and the problem of insufficient data volume is solved; effective management of multiple servers is achieved through a Hadoop platform, and the problem that mass data are difficult to collect and store is solved; the target detection algorithm can effectively and quickly identify plant diseases; a model training task platform simplifies a model training task process; and a visual display platform provides information service for farmers in a visual and easy-to-understand manner. The platform can assist farmers in timely and effective recognition of diseases and insect pests, and economic loss caused by spreading of diseases and insect pests is prevented.

Description

technical field [0001] The invention relates to the field of big data and deep learning, in particular to a pest identification system based on big data and deep learning. Background technique [0002] Correlation diagnosis based on diseased plant images is one of the research hotspots of smart agriculture. It is necessary to distinguish the lesion traits (color, area, ulceration) of the diseased image and finally get the diseased type, which has high complexity. Because of this complexity, even experienced agronomists and plant pathologists are often unable to successfully diagnose a particular disease, leading to erroneous conclusions and treatments; As well as new technologies of artificial intelligence, image processing and graphics processing unit (Graphics Processing Unit, GPU) have made remarkable progress, which has promoted the rapid development of artificial intelligence, which can improve the accuracy of identifying pests and diseases. For farmers who lack appropr...

AI Technical Summary

Problems solved by technology

[0003] In the pest recognition system, many pest data classification tasks are involved. In order to obtain higher classification accuracy, the number of pest pictures is a key factor, so the data of pest pictures has high research and utilization value; When it comes to image data, the traditional single-node crawler technology has a slow crawling speed for data with a large amount of data and many types of data, and it is difficult to obtain a large amount of data required for model training; among them, the process of model training is cumbersome, and model update is troublesome. It is easy to cause idle servers and waste of computing power; in addition, traditional technology deployment is inconvenient, deployment versions are chaotic, deployment isolation is poor, and scripts and codes are difficult to manage

Images