Accurate and early detection of poplar tree leaf spot disease using image processing technique.

Leaf spot disease is one of the most common fungal diseases that cause immense and sometimes irreparable damage to poplar trees. In this regard, conventional methods for detection of fungal contamination are time-consuming, costly and difficult. Therefore, in this study, in order to distinguish healthy leaves from infected ones, a modified image processing algorithm by using the Laplacian threshold was used. In this study, the illumination wavelengths have been selected using a UV-Visible Spectrophotometer device (PG Instruments, Model T92+ Double® Beam, England). The images were recorded using a 1260 lux light intensity and a digital camera mounted on the stereomicroscope (SN 287050, Italy) with a resolution of 5 megapixels. The experimental data were analyzed based on a completely randomized design with three replications. Analysis of variance (ANOVA) test was used to determine the significance of treatments performance. According to the results, the red spectrum with 680nm wavelength showed the highest contaminated surface on the leaf (190500 μm²). In contrast, the yellow spectrum with a wavelength of 585 nm determined the lowest amount of contaminated surface (65781 μm²). The blue and green spectra showed roughly the same performance in early detection of fungal contamination. Overall obtained results showed that the red spectrum with wavelength of 680 nm is more reliable for early poplar leaves’ surface contamination detection in compare to blue (470 nm), green (550 nm) and yellow (585 nm) wavelengths with 13%, 14% and 55.8% improvements, respectively. The method presented in this study can be used to identify the quality and the health of biological products and disease progression, significantly easier and faster in compare to the conventional methods. [ABSTRACT FROM AUTHOR]