Open Access

ARTICLE

Research on Prevention of Citrus Anthracnose Based on Image Retrieval Technology

Xuefei Du^*, Xuyu Xiang

College of Computer and Information Engineering, Central South University of Forestry and Technology, Changsha, China

* Corresponding Author: Xuefei Du. Email:

Journal of Information Hiding and Privacy Protection 2020, 2(1), 11-19. https://doi.org/10.32604/jihpp.2020.010114

Received 10 May 2020; Accepted 28 May 2020; Issue published 15 October 2020

Abstract

Citrus anthracnose is a common fungal disease in citrus-growing areas in China, which causes very serious damage. At present, the manual management method is time-consuming and labor-consuming, which reduces the control effect of citrus anthracnose. Therefore, by designing and running the image retrieval system of citrus anthracnose, the automatic recognition and analysis of citrus anthracnose control were realized, and the control effect of citrus anthracnose was improved. In this paper, based on the self-collected and collated citrus anthracnose image database, we use three image features to realize an image retrieval system based on citrus anthracnose through SMV, AP clustering optimization. The results show that: 1) In the accuracy of image feature retrieval, Gist feature extraction > cumulative color histogram > Gabor texture feature; 2) In the maximum divergence diversity retrieval, semi-supervised AP clustering retrieval > AP clustering retrieval > SVM relevance feedback results > initial retrieval. 3) Practice shows that this technology can reduce the workload of monitoring and management in the control process of citrus planting area, and promote the intelligent and efficient control of citrus anthracnose, which has high practical application value.

---

Keywords

---