Hyperspectral Imaging for Seed Viability: A Review

Authors

  • Robin S. Athawale Department of Computer Science and Information Technology, Dr. Babasaheb Ambedkar Marathwada University, India https://orcid.org/0009-0007-6278-7303
  • Ramdas D. Gore Department of Computer Science and Information Technology, Dr. Babasaheb Ambedkar Marathwada University, India https://orcid.org/0000-0002-9703-405X
  • Bharti W. Gawali Department of Computer Science and Information Technology, Dr. Babasaheb Ambedkar Marathwada University, India https://orcid.org/0000-0002-8353-5849

DOI:

https://doi.org/10.47852/bonviewJDSIS3202836

Keywords:

Pika L, spectral signature, PCA, PLSR, SVM, CNN, YOLOv5

Abstract

Seed is the largest crop in the world and is a very usable crop in agriculture and widely cultivated. Seed grain is a worldwide essential food. It is the evolution of agriculture, and cultivating seeds for harvest was about 9000 to 10,000 years ago. The scientific name of wheat seed is “Triticum.” There are two types of seeds: monocotyledonous and dicotyledonous. Wheat seeds consist of a monocotyledonous kind. In this review, we have studied seed viability. Viability means quality or health or state of being viable. Seed viability plays a very vital role in agriculture. We have also studied the hyperspectral imaging tool. This paper overviews seed viability, hyperspectral imaging technology, spectral signature, and some preprocessing and classification techniques such as noise removal, image filter, Savitzky–Golay (SG), multiplicative scatter correction (MSC), and standard normal variate (SNV). A classification technique is used for separating various spectral classes. The seed viability used some testing such as a control test, cool test, warm test, Petri paper dish test, and chemical test (tetrazolium (TZ) test) to check seed viability. This paper helps in choosing the two ways of testing seed viability. One is using hyperspectral imaging and analyzing techniques on the computer, and the second is biological techniques. A comparative study of different technique is done as per stages. The seed parameters are shape, size, weight, and chemical components for analyzing the seed viability using hyperspectral imaging. The hyperspectral imaging wavelength range is between 400 and 1000 nm. This paper overviews biological testing and hyperspectral imaging tool for seed variability.

 

Received: 8 March 2023 | Revised: 19 April 2023 | Accepted: 10 May 2023

 

Conflicts of Interest

The authors declare that they have no conflicts of interest to this work.

 

Data Availability Statement

Data sharing is not applicable to this article as no new data were created or analyzed in this study


Downloads

Published

2023-05-29

How to Cite

Athawale, R. S., Gore, R. D., & Gawali, B. W. (2023). Hyperspectral Imaging for Seed Viability: A Review. Journal of Data Science and Intelligent Systems. https://doi.org/10.47852/bonviewJDSIS3202836

Issue

Section

Review