C.L. Kim, C. Li, A. Raheja, and D.W. Still (USA)
Seed germination, object detection, Hough transform, noise reduction, object labelling, image segmentation
This paper presents a complete process of extracting seed characteristics used for germination detection by using image processing related to circle detection, noise filtering, image segmentation, thresholding, morphological operations, and object labeling. Implementations of these tools have been optimized to best fit the purpose of this system. The purpose of the presented system is to allow seed testing, which measures the quality of the seeds using germination detection, to be more rapid, objective, and simple. This is important because planting high quality seeds have been the basis by which agriculture remains profitable. Every seed lot sold commercially has had an assessment of seed quality which requires seeds to germinate under multiple environmental conditions. This paper describes an approach to automate seed analysis and reduce a bottleneck that exists in evaluating seed germination data which would otherwise be evaluated by humans. The proposed procedure detects a given Petri dish, separates the background from the seeds, and computes necessary seed parameters used for analysis.
Important Links:
Go Back
