Assessment of paddy leaves disease severity level using image processing technique

Abstract

This study aims to develop an automated image acquisition system for paddy leaves images and an image processing algorithm to estimate the severity level of the disease. The images were acquired using the developed system and processed using MATLAB software. By quantifying the total and infected leaf area based on pixel counting, the algorithm calculated the percentage of the infected leaf area using a ratio-based method. Through the image processing method, the study achieved a range of 15.53% to 41.23% of the total percentage of infected leaf area. In comparison, the leaf area meter method yielded a lower range of 23.49% to 28.98% of the total infected area. This indicates that the image-processing approach can capture a wider extent of infection. However, a deviation was observed between the two methods, ranging from -9.39% to 17.74% in the image processing results. Ensuring proper lighting conditions and installing an automatic camera movement system would significantly enhance the accuracy of disease severity measurements.

Keywords

• Paddy leaf diseases
• Disease quantification
• Image processing
• Disease severity