Detection method of rice blast based on 4D light field refocusing depth information fusion.

Rice blast is one of the most widely distributed rice diseases in the world. Its rapid transmission and strong harm have seriously affected the rice yield. The traditional detection methods for rice blast are time-consuming and laborious. Therefore, rapid and accurate detection of rice blast in its early stage has become a priority. In this study, a rice blast detection method based on 4D light field refocusing depth information fusion is studied, which realizes the decoding of rice blast light field images, light field refocusing and image detection after refocusing. The four-dimensional light information recorded by the light field is used to quickly lock the best focal plane of the rice blast image. Then analyze the characteristics of rice blast and carry out subsequent image processing. The experimental results show that the average accuracy of the method used in the study is 96.08%, which has a slight advantage over other image detection methods. The average detection time is 12.13 s, which is a big improvement compared to the manual detection method (38.91 s) and the traditional camera-based detection method (16.68 s). The rice blast detection method based on 4D light field refocusing depth information fusion has certain advantages compared with other image detection and artificial detection methods and can be used for the detection of rice blast.