A halation reduction method for high quality images of tomato fruits in greenhouse

Abstract

Halation is a serious problem for machine vision systems which can cause loss of color information from an image. Images of tomato fruits grown in a greenhouse taken with a different rotational angle polarizing filter (PL filter) were acquired and combined into one resulting image to eliminate multiple halation areas due to sunlight and reflected light. An image acquisition system composed of a CCD camera and a stepping motor was used to acquire the images. The multiple halation areas could be eliminated at a certain angle of incidence and the color representation of the resultant image was improved. The results demonstrate the feasibility of eliminating multiple halation areas using this method.