Research on automatic identification system of tobacco diseases

ABSTRACT

In order to improve recognition accuracy of tobacco diseases, an identification method based on multi-feature and genetic algorithms optimizing BP neural network was proposed. First, Otsu method was used to obtain disease location information and GrabCut function was initialized for extracting diseased area effectively. Second, colour moments, disease contour and GLCM were used to get colour, multi-contour and texture features. Once again, BP neural network was optimized by genetic algorithm, and the optimal initial weights and thresholds were obtained, which shortened the training time and improved the accuracy of disease identification. Finally, BP neural network model for tobacco diseases diagnosis was established with the mobile client as input and the user services as output. The field experiment showed that the method could diagnose eight types of tobacco diseases effectively and automatically. The average recognition accuracy rate of selected tobacco diseases was about 92.5%.

KEYWORDS:

• Tobacco disease
• interactive segmentation
• multi-feature
• genetic algorithm
• BP neural network

Notes on contributors

Yuanyuan Shao is an associate professor in Mechanical and Electronic Engineering College, Shandong Agricultural University. Yuanyuan Shao obtained PhD in Mechanical and Electronic Engineering, and focuses on agricultural mechanisation and image processing.

Guantao Xuan is a faculty in Mechanical and Electronic Engineering College, Shandong Agricultural University. Guantao Xuan obtained PhD in Agricultural Mechanisation Engineering, and focuses on precision agriculture.

Yangyan Zhu is an undergraduate student in Electric Engineering and Automation Department, Shandong Agricultural University.

Yanling Zhang is a postgraduate student in Electric Engineering and Automation Department, Shandong Agricultural University, and focuses on image processing.

Hongxing Peng is an associate professor in Computer Science and Engineering Department in South China Agricultural University, and focuses on computer vision, image processing, robot, virtual reality and human–computer interaction.

Zhongzheng Liu obtained masters in Agricultural Mechanisation Engineering from Shandong Agricultural University, and focuses on agricultural mechanisation.

Jialin Hou is a professor and dean of Mechanical and Electronic Engineering College, Shandong Agricultural University, and focuses on automation of agricultural machinery.

Additional information

Funding

This work is supported by Modern Agricultural Technology System of Shandong Province (Grant No. SDAIT-02-021-09), ‘National Natural Science Foundation of China (Grant No. 31671632)’, ‘Key research and development program of Shandong Province (Grant No. 2016GNC112007)’, ‘Natural Science Foundation of Shandong Province (Grant No. BS2015NY012)’, and also supported by ‘Natural Science Foundation of Guangdong Province (2015A030310258)’.