Denmark
October 28, 2021
The journal of Sensors has published a new paper showing how VideometerLab images are used to discriminate between seed varieties. The research depicts how spectral imaging combined with machine learning can help scientists with determining eggplant seed attributes.
A New Paper on Seed Variety Discrimination With the VideometerLab
In the experiment seventeen eggplant seed varieties were used, which were cultivated in Hebei, a Chinese province. The authors of the paper used the VideometerLab 4 to capture images of mixed seed varieties. The classification model was prepared with the assistance of MatLab. The further analysis together with image processing was conducted through the VideometerLab software, here they were able to visualize chemical and physical features of each seed variety. The author’s state:
Seventy-eight features acquired with the multispectral images were extracted from individual eggplant seeds, which were then classified using SVM and a one-dimensional convolutional neural network (1D-CNN), and the overall accuracy was 90.12% and 94.80%, respectively. A two-dimensional convolutional neural network (2D-CNN) was also adopted for discrimination of seed varieties, and an accuracy of 90.67% was achieved.
Lei Sun, Xiaofei Fan, Sheng Huang, Shuangxia Luo, Lili Zhao, Xueping Chen, Yi He, Xuesong Suo, “Research on Classification Method of Eggplant Seeds Based on Machine Learning and Multispectral Imaging Classification Eggplant Seeds”, Journal of Sensors, vol. 2021, Article ID 8857931, 9 pages, 2021.
Genetic purity
Seed variety determination – also known as genetic purity – is a critical tool for seed producers and seed breeders since inadvertent mixing of seeds and seed varieties may compromise seed quality. Furthermore, the discrimination of seed traits can help ensure crop quality.
Certified seeds are seeds with 99% of genetic purity, usually used for commercial purposes, and produced by private companies or state-owned corporations. Those seeds are inspected under strict certification standards to ensure varietal purity. (ag.ndsu.edu)
For many years the manual inspection was the most common way of examining the seed varieties. However, this method is time-consuming and not so accurate.
Videometer proposes their state-of-art technology for seed analysis. Spectral imaging can assist with assessing seed purity, health, germination, and more. VideometerLab can identify crucial information which is not visible to the naked eye. The automated analysis is powerful, fast, objective, and non-destructive.
Explore Videometer’s seed applications by clicking on the button below!
Seed Applications
References
Lei Sun, Xiaofei Fan, Sheng Huang, Shuangxia Luo, Lili Zhao, Xueping Chen, Yi He, Xuesong Suo, “Research on Classification Method of Eggplant Seeds Based on Machine Learning and Multispectral Imaging Classification Eggplant Seeds”, Journal of Sensors, vol. 2021, Article ID 8857931, 9 pages, 2021.
Topics
Species
