Abstract

Distinctiveness, Uniqueness and Stability (DUS) testing is a standard practice to establish a new variety of crop. Color measurement plays a major role in DUS testing. Color of plant parts like leaf, flower, fruit and stem are very important factors to establish the uniqueness of any candidate variety. To measure the color or to describe a color objectively by some name or code, Royal Horticultural Society (RHS) color chart is used by the DUS tester. RHS color chart is available in market. In manual process of color measurement, samples are placed under the porthole of the selected color-field of the RHS color chart to find the best matched color based on the visual perception of the person being. Visual evaluation is a tricky process as it depends on so many factors like lighting condition of the environment, gender, age, eyesight of the evaluator etc. Also, selecting of the proper color-field from the color fan of RHS color chart is fishy and it may lead to a biased result. In this paper, it is proposed to use an intelligent machine vision technology which will accurately and objectively measure color of plant parts in quick time. A database of colorimetric values of all RHS colors was used as a reference. Digital image processing and analysis algorithms were applied on the images of plant part to extract the color information using RGB, CIELab and CIELCh color models. Extracted color information was compared with the available color values in the reference database of RHS color chart using Euclidean distance, Chebyshev distance and Mahalanobis distance. These distance values were used to determine the top five matching color for each color model. Top priority was given to the color with minimum distance value. It was observed that CIELab color model using Mahalonobis distance was the most excellent model for color comparison while DUS testing. This technology may definitely help the DUS tester to accurately describe the color of plant parts in quick time.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call