Rapid Detection of Rice Disease Based on FCM-KM and Faster R-CNN Fusion

Guoxiong Zhou,Aibin Chen,Wenzhuo Zhang,Mingfang He,Xueshuo Ma

doi:10.1109/access.2019.2943454

Guoxiong Zhou, Aibin Chen + Show 3 more

Open Access

https://doi.org/10.1109/access.2019.2943454

Copy DOI

Abstract

In this paper, a method for detecting rapid rice disease based on FCM-KM and Faster R-CNN fusion is proposed to address various problems with the rice disease images, such as noise, blurred image edge, large background interference and low detection accuracy. Firstly, the method uses a two-dimensional filtering mask combined with a weighted multilevel median filter (2DFM-AMMF) for noise reduction, and uses a faster two-dimensional Otsu threshold segmentation algorithm (Faster 2D-Otsu) to reduce the interference of complex background with the detection of target blade in the image. Then the dynamic population firefly algorithm based on the chaos theory as well as the maximum and minimum distance algorithm is applied for optimization of the K-Means clustering algorithm (FCM-KM) to determine the optimal clustering class k value while addressing the tendency of the algorithm to fall into the local optimum problem. Combined with the R-CNN algorithm for the identification of rice diseases, FCM-KM analysis is conducted to determine the different sizes of the Faster R-CNN target frame. As revealed by the application results of 3010 images, the accuracy and time required for detection of rice blast, bacterial blight and blight were 96.71%/0.65s, 97.53%/0.82s and 98.26%/0.53s, respectively, indicating clearly that the method is more capable of detecting rice diseases and improving the identification accuracy of Faster R-CNN algorithm, while reducing the time required for identification.

Highlights

As one of world’s major food crops, rice is stable in production, which is related to agricultural security, social stability and national development
We have proposed a method for detecting rapid rice disease based on FCM-KM and fast R-CNN fusion
TEST RESULTS 1) RICE BLAST For the rice blast, the recognition results are shown in Figure 11, which indicates that the method put forward in this paper can accurately identify rice blast lesions

Summary

Introduction

As one of world’s major food crops, rice is stable in production, which is related to agricultural security, social stability and national development. The identification of rice diseases is mainly through artificial identification, querying rice diseases maps and automated detection. CNN The team of the target detection community, Ross Girshick, launched a new effort Faster R-CNN [21] in 2015, which was after the launch of R-CNN [29] and Fast R-CNN [30]. Faster R-CNN does not have a fixed size requirement for the image of rice diseases to be detected. Faster R-CNN can be viewed as a model of ‘‘the regional generation network + Fast R-CNN’’, which applies the Region Proposal. Network (RPN) instead of Selective Search in Fast R-CNN. The convolution layer/full connection layer processing is performed on the feature map, and the detected target is subjected to position regression and the classification, and the region recommendation is used to obtain a more accurate disease location. Fast R-CNN is in line with the detailed calculation of the position of the frame and the category of the objects in the frame

Methods

Results

Conclusion