TY - GEN
T1 - Portable detection system of vegetable oils based on Laser induced fluorescence
AU - Zhu, Li
AU - Zhang, Yinchao
AU - Chen, Siying
AU - Chen, He
AU - Guo, Pan
AU - Mu, Taotao
N1 - Publisher Copyright:
© 2015 SPIE.
PY - 2015
Y1 - 2015
N2 - Food safety, especially edible oils, has attracted more and more attention recently. Many methods and instruments have emerged to detect the edible oils, which include oils classification and adulteration. It is well known than the adulteration is based on classification. Then, in this paper, a portable detection system, based on laser induced fluorescence, is proposed and designed to classify the various edible oils, including (olive, rapeseed, walnut, peanut, linseed, sunflower, corn oils). 532 nm laser modules are used in this equipment. Then, all the components are assembled into a module (100∗100∗25mm). A total of 700 sets of fluorescence data (100 sets of each type oil) are collected. In order to classify different edible oils, principle components analysis and support vector machine have been employed in the data analysis. The training set consisted of 560 sets of data (80 sets of each oil) and the test set consisted of 140 sets of data (20 sets of each oil). The recognition rate is up to 99%, which demonstrates the reliability of this potable system. With nonintrusive and no sample preparation characteristic, the potable system can be effectively applied for food detection.
AB - Food safety, especially edible oils, has attracted more and more attention recently. Many methods and instruments have emerged to detect the edible oils, which include oils classification and adulteration. It is well known than the adulteration is based on classification. Then, in this paper, a portable detection system, based on laser induced fluorescence, is proposed and designed to classify the various edible oils, including (olive, rapeseed, walnut, peanut, linseed, sunflower, corn oils). 532 nm laser modules are used in this equipment. Then, all the components are assembled into a module (100∗100∗25mm). A total of 700 sets of fluorescence data (100 sets of each type oil) are collected. In order to classify different edible oils, principle components analysis and support vector machine have been employed in the data analysis. The training set consisted of 560 sets of data (80 sets of each oil) and the test set consisted of 140 sets of data (20 sets of each oil). The recognition rate is up to 99%, which demonstrates the reliability of this potable system. With nonintrusive and no sample preparation characteristic, the potable system can be effectively applied for food detection.
KW - Classification
KW - Edible oils
KW - Laser induced fluorescence
KW - Portable
KW - Priciple components analysis
KW - Support vector machine
UR - http://www.scopus.com/inward/record.url?scp=84963744354&partnerID=8YFLogxK
U2 - 10.1117/12.2217577
DO - 10.1117/12.2217577
M3 - Conference contribution
AN - SCOPUS:84963744354
T3 - Proceedings of SPIE - The International Society for Optical Engineering
BT - Selected Papers of the Photoelectronic Technology Committee Conferences held June-July 2015
A2 - Lv, Daren
A2 - Wang, Lijun
A2 - Xiang, Libin
A2 - Petelin, Michael I.
A2 - Zhang, Guangjin
A2 - Zhuang, Songlin
A2 - Liu, Shenggang
PB - SPIE
T2 - International Conference on Frontiers in International Conference on Frontiers in Terahertz Technology and Applications, and the International Symposium on Surface Topography and Optical Microscopy
Y2 - 23 July 2015 through 25 July 2015
ER -