Skip to main content
Fig. 2 | Plant Methods

Fig. 2

From: A comparison of ImageJ and machine learning based image analysis methods to measure cassava bacterial blight disease severity

Fig. 2

Manual ImageJ analysis of CBB water-soaking symptoms. A Images of cassava leaves infiltrated with Xam WT, XamΔTAL20, and mock treatments were segmented and analyzed using an ImageJ overlay segmentation method. Overlay segmentation analysis depicted by step using a CBB infected cassava leaf image. Images were taken at 0, 4, 6 and 9 DPI. Leaf lobes were labeled by treatment type: X = Xam WT, T = XamΔTAL20, and M = Mock. White lines point to selected regions of a representative water-soaked lesion at each step of the ImageJ overlay segmentation process. B The variance explained by inoculation type (Xam WT or XamΔTAL20) DPI (4-, 6- and 9-), or the interaction between inoculation type and DPI for ten ImageJ generated measurements. Variances were determined by ANOVA analysis. C Total water-soaked area (pixels, y-axis) for sites infiltrated with each treatment (x-axis). Calculated p-values (Kolmogorov–Smirnov test) shown above the line in each plot. D Negative gray-scale mean (y-axis) of water-soaked lesions for Xam WT and XamΔTAL20 relative to mock inoculated spots (x-axis) within the same leaf. Calculated p-values (Kolmogorov–Smirnov test) shown above the line in each plot. In ImageJ, the gray-scale mean was measured by averaging the mean of each gray-scale value in the RGB channels

Back to article page