Assessing mango anthracnose using a new three-dimensional image-analysis technique to quantify lesions on fruit

Abstract

An accurate image-analysis method was developed to assess quantitatively the spot-like lesions on fruits resulting from pathogen attack. The technique was applied to evaluation of the development and severity of anthracnose of mango fruit, caused by the fungus Colletotrichum gloeosporioides. In this method, a stepper motor rotates the mango fruit along its longitudinal axis while acquiring a sequence of 360 images of its total surface (one image for each degree). This set of images is used to create a pseudocylindrical 'equal-area' projection of the fruit in a two-dimensional map containing complete morphometrical and photometrical information of its surface. The lesion area can easily be evaluated from this map with image-analysis procedures. Quantitative data (percentage of area affected) can be used to establish an assessment scale for the disease based on lesion spots measured, as well as for detailed laboratory studies of mango anthracnose development. The average error of the method is -0·1%, standard deviation 0·44 (r 2 = 0·99), and it may be adapted for use with most commercial image analysers and for other diseases with spot-like symptoms. © 2005 BSPP.