Search

The spread of citrus canker has become one of the most important challenges faced by Florida Fresh Market citrus is affecting the export of citrus fruits to several international markets including European countries. Previous studies have shown that automated detection systems can help detect citrus canker infected fruit and could assist in eliminating the detected fruit from shipment to closed markets. Most automated detection systems use some form of machine vision with artificial light sources. However, when capturing images of spherical objects, non-uniform illumination results in an edge blackening effect resulting in higher misclassification rate. The basic objective of this research was to implement a compensation algorithm to eliminate the edge effect when capturing hyperspectral image of citrus fruits. The algorithm originally developed by Gomez et al. 2007, was adapted for the purpose of canker detection. The image was corrected for spatial variations (flat field correction) caused by intensity of light source as well as geometrical variation caused by the spherical geometry of the citrus fruit. In this study, the geometric correction was accomplished by constructing a 3-D digital elevation model (DEM) of the fruit from its 2-D image. This DEM provided the geometric properties of the fruit X, Y, and Z coordinates which were exploited in the course of estimating the geometric correction factor for each pixel. The corrected image portrayed a more uniform brightness of the citrus fruit surface throughout. Tests were conducted on 10 orange samples (five marketable and five cankerous) to validate the results of the algorithm which demonstrated that the geometric correction resulted in uniform intensity of radiation throughout the fruit surface thus reducing the within class variation. [ABSTRACT FROM AUTHOR]