Rice seed varietal purity inspection using hyperspectral imaging

Research output: Contribution to conferenceAbstract

Abstract

When distributing rice seed to farmers, suppliers strive to ensure that all seeds delivered belong to the species that was ordered and that the batch is not contaminated by unhealthy seeds or seeds of a different species. A conventional method to inspect the varietal purity of rice seeds is based on manually selecting random samples of rice seed from a batch and evaluating the physical grain properties through a process of human visual inspection. This is a tedious, laborious, time consuming and extremely inefficient task where only a very small subset of the entire batch of the rice seed can be examined. There is, therefore, a need to automate this process to make it repeatable and more efficient while allowing a larger sample of rice seeds from any batch to be analysed. This paper presents an automatic rice seed inspection method which combines hyperspectral imaging and tools from machine learning to automatically detect seeds which are erroneously contained within a batch when they actually belong to a completely different species. Image data from Near-infrared (NIR) and Visible Light (VIS) hyperspectral cameras are acquired for six common rice seed varieties. Two different classifiers are applied to the data: a Support Vector Machine (SVM) and a Random Forest (RF), where each consists of six one-versus-rest binary classifiers. The results show that combining spectral and shape-based features derived from the rice seeds results in an increase in the precision (PPV) of the multi-label classification to 84% compared with 74% when only visual features are used.

Conference

ConferenceHyperspectral Imaging and Applications Conference
Abbreviated titleHSI 2016
CountryUnited Kingdom
CityCoventry
Period12/10/1613/10/16

Fingerprint

Seed
Inspection
Hyperspectral imaging
Classifiers
Support vector machines
Learning systems
Labels
Cameras
Infrared radiation

Keywords

  • hyperspectral imaging
  • rice seed
  • automation

Cite this

Hai, V., Tachtatzis, C., Murray, P., Harle, D., Dao, T. K., Le, T-L., ... Marshall, S. (2016). Rice seed varietal purity inspection using hyperspectral imaging. Abstract from Hyperspectral Imaging and Applications Conference, Coventry, United Kingdom.
Hai, Vu ; Tachtatzis, Christos ; Murray, Paul ; Harle, David ; Dao, Trung Kien ; Le, Thi-Lan ; Andonovic, Ivan ; Marshall, Stephen. / Rice seed varietal purity inspection using hyperspectral imaging. Abstract from Hyperspectral Imaging and Applications Conference, Coventry, United Kingdom.
@conference{588aceb7a67b4b51969de1b0fc10635b,
title = "Rice seed varietal purity inspection using hyperspectral imaging",
abstract = "When distributing rice seed to farmers, suppliers strive to ensure that all seeds delivered belong to the species that was ordered and that the batch is not contaminated by unhealthy seeds or seeds of a different species. A conventional method to inspect the varietal purity of rice seeds is based on manually selecting random samples of rice seed from a batch and evaluating the physical grain properties through a process of human visual inspection. This is a tedious, laborious, time consuming and extremely inefficient task where only a very small subset of the entire batch of the rice seed can be examined. There is, therefore, a need to automate this process to make it repeatable and more efficient while allowing a larger sample of rice seeds from any batch to be analysed. This paper presents an automatic rice seed inspection method which combines hyperspectral imaging and tools from machine learning to automatically detect seeds which are erroneously contained within a batch when they actually belong to a completely different species. Image data from Near-infrared (NIR) and Visible Light (VIS) hyperspectral cameras are acquired for six common rice seed varieties. Two different classifiers are applied to the data: a Support Vector Machine (SVM) and a Random Forest (RF), where each consists of six one-versus-rest binary classifiers. The results show that combining spectral and shape-based features derived from the rice seeds results in an increase in the precision (PPV) of the multi-label classification to 84{\%} compared with 74{\%} when only visual features are used.",
keywords = "hyperspectral imaging, rice seed, automation",
author = "Vu Hai and Christos Tachtatzis and Paul Murray and David Harle and Dao, {Trung Kien} and Thi-Lan Le and Ivan Andonovic and Stephen Marshall",
year = "2016",
month = "10",
day = "13",
language = "English",
note = "Hyperspectral Imaging and Applications Conference, HSI 2016 ; Conference date: 12-10-2016 Through 13-10-2016",

}

Hai, V, Tachtatzis, C, Murray, P, Harle, D, Dao, TK, Le, T-L, Andonovic, I & Marshall, S 2016, 'Rice seed varietal purity inspection using hyperspectral imaging' Hyperspectral Imaging and Applications Conference, Coventry, United Kingdom, 12/10/16 - 13/10/16, .

Rice seed varietal purity inspection using hyperspectral imaging. / Hai, Vu; Tachtatzis, Christos; Murray, Paul; Harle, David; Dao, Trung Kien; Le, Thi-Lan; Andonovic, Ivan; Marshall, Stephen.

2016. Abstract from Hyperspectral Imaging and Applications Conference, Coventry, United Kingdom.

Research output: Contribution to conferenceAbstract

TY - CONF

T1 - Rice seed varietal purity inspection using hyperspectral imaging

AU - Hai, Vu

AU - Tachtatzis, Christos

AU - Murray, Paul

AU - Harle, David

AU - Dao, Trung Kien

AU - Le, Thi-Lan

AU - Andonovic, Ivan

AU - Marshall, Stephen

PY - 2016/10/13

Y1 - 2016/10/13

N2 - When distributing rice seed to farmers, suppliers strive to ensure that all seeds delivered belong to the species that was ordered and that the batch is not contaminated by unhealthy seeds or seeds of a different species. A conventional method to inspect the varietal purity of rice seeds is based on manually selecting random samples of rice seed from a batch and evaluating the physical grain properties through a process of human visual inspection. This is a tedious, laborious, time consuming and extremely inefficient task where only a very small subset of the entire batch of the rice seed can be examined. There is, therefore, a need to automate this process to make it repeatable and more efficient while allowing a larger sample of rice seeds from any batch to be analysed. This paper presents an automatic rice seed inspection method which combines hyperspectral imaging and tools from machine learning to automatically detect seeds which are erroneously contained within a batch when they actually belong to a completely different species. Image data from Near-infrared (NIR) and Visible Light (VIS) hyperspectral cameras are acquired for six common rice seed varieties. Two different classifiers are applied to the data: a Support Vector Machine (SVM) and a Random Forest (RF), where each consists of six one-versus-rest binary classifiers. The results show that combining spectral and shape-based features derived from the rice seeds results in an increase in the precision (PPV) of the multi-label classification to 84% compared with 74% when only visual features are used.

AB - When distributing rice seed to farmers, suppliers strive to ensure that all seeds delivered belong to the species that was ordered and that the batch is not contaminated by unhealthy seeds or seeds of a different species. A conventional method to inspect the varietal purity of rice seeds is based on manually selecting random samples of rice seed from a batch and evaluating the physical grain properties through a process of human visual inspection. This is a tedious, laborious, time consuming and extremely inefficient task where only a very small subset of the entire batch of the rice seed can be examined. There is, therefore, a need to automate this process to make it repeatable and more efficient while allowing a larger sample of rice seeds from any batch to be analysed. This paper presents an automatic rice seed inspection method which combines hyperspectral imaging and tools from machine learning to automatically detect seeds which are erroneously contained within a batch when they actually belong to a completely different species. Image data from Near-infrared (NIR) and Visible Light (VIS) hyperspectral cameras are acquired for six common rice seed varieties. Two different classifiers are applied to the data: a Support Vector Machine (SVM) and a Random Forest (RF), where each consists of six one-versus-rest binary classifiers. The results show that combining spectral and shape-based features derived from the rice seeds results in an increase in the precision (PPV) of the multi-label classification to 84% compared with 74% when only visual features are used.

KW - hyperspectral imaging

KW - rice seed

KW - automation

UR - http://www.hsi2016.com/

M3 - Abstract

ER -

Hai V, Tachtatzis C, Murray P, Harle D, Dao TK, Le T-L et al. Rice seed varietal purity inspection using hyperspectral imaging. 2016. Abstract from Hyperspectral Imaging and Applications Conference, Coventry, United Kingdom.