Comparison of color model for flower recognition

Perani Rosyani; M Taufik; Arya Adhyaksa Waskita; Diah Harnoni  Apriyanti

doi:10.1109/ICITISEE.2018.8721026

Comparison of color model for flower recognition

Perani Rosyani, M Taufik, Arya Adhyaksa Waskita, Diah Harnoni Apriyanti

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

Abstract

Comparison to RGB, HSV, LAB and YCbCr color model in a flower recognition has been conducted for 12 species from two families. Flower images were obtained from the ImageCLEF 2017 dataset. After segmenting the flower from its background using the k-means clustering method, the statistical parameters are extracted from each color model. These parameters including the minimum and maximum pixel value, and also mean and standard deviation. Then, the classifying result of SVM shows that the HSV color model performs the worst amongst the other investigated color model, which is around 30% using different kernel. While the other color model has the accuracy around 70% using linear and polynomial kernel. This result in line with the surface plot of their statistical characteristics.

Original language	English
Title of host publication	3rd International Conference on Information Technology, Information System and Electrical Engineering 2018
Publisher	IEEE
Pages	10-15
Number of pages	6
ISBN (Electronic)	978-1-5386-7082-8
ISBN (Print)	978-1-5386-7083-5
DOIs	https://doi.org/10.1109/ICITISEE.2018.8721026
Publication status	Published - 23 May 2019
Externally published	Yes
Event	3rd International Conference on Information Technology, Information System and Electrical Engineering 2018 - Grand Inna Malioboro Hotel, Yogyakarta, Indonesia Duration: 13 Nov 2018 → 14 Nov 2018 Conference number: 3 http://icitisee.amikompurwokerto.ac.id/

Conference

Conference	3rd International Conference on Information Technology, Information System and Electrical Engineering 2018
Abbreviated title	ICITISEE 2018
Country	Indonesia
City	Yogyakarta
Period	13/11/18 → 14/11/18
Internet address	http://icitisee.amikompurwokerto.ac.id/

Keywords

image colour analysis
feature extraction
support vector machines

Access to Document

10.1109/ICITISEE.2018.8721026

Fingerprint Dive into the research topics of 'Comparison of color model for flower recognition'. Together they form a unique fingerprint.

Cite this

Rosyani, P., Taufik, M., Waskita, A. A., & Apriyanti, D. H. (2019). Comparison of color model for flower recognition. In 3rd International Conference on Information Technology, Information System and Electrical Engineering 2018 (pp. 10-15). IEEE. https://doi.org/10.1109/ICITISEE.2018.8721026

@inproceedings{4289eabf19ae4163b8d05d638a635033,

title = "Comparison of color model for flower recognition",

abstract = "Comparison to RGB, HSV, LAB and YCbCr color model in a flower recognition has been conducted for 12 species from two families. Flower images were obtained from the ImageCLEF 2017 dataset. After segmenting the flower from its background using the k-means clustering method, the statistical parameters are extracted from each color model. These parameters including the minimum and maximum pixel value, and also mean and standard deviation. Then, the classifying result of SVM shows that the HSV color model performs the worst amongst the other investigated color model, which is around 30{\%} using different kernel. While the other color model has the accuracy around 70{\%} using linear and polynomial kernel. This result in line with the surface plot of their statistical characteristics.",

keywords = "image colour analysis, feature extraction, support vector machines",

author = "Perani Rosyani and M Taufik and Waskita, {Arya Adhyaksa} and Apriyanti, {Diah Harnoni}",

year = "2019",

month = "5",

day = "23",

doi = "10.1109/ICITISEE.2018.8721026",

language = "English",

isbn = "978-1-5386-7083-5",

pages = "10--15",

booktitle = "3rd International Conference on Information Technology, Information System and Electrical Engineering 2018",

publisher = "IEEE",

address = "United States",

}

Rosyani, P, Taufik, M, Waskita, AA & Apriyanti, DH 2019, Comparison of color model for flower recognition. in 3rd International Conference on Information Technology, Information System and Electrical Engineering 2018. IEEE, pp. 10-15, 3rd International Conference on Information Technology, Information System and Electrical Engineering 2018, Yogyakarta, Indonesia, 13/11/18. https://doi.org/10.1109/ICITISEE.2018.8721026

Comparison of color model for flower recognition. / Rosyani, Perani; Taufik, M; Waskita, Arya Adhyaksa; Apriyanti, Diah Harnoni .

3rd International Conference on Information Technology, Information System and Electrical Engineering 2018. IEEE, 2019. p. 10-15.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - Comparison of color model for flower recognition

AU - Rosyani, Perani

AU - Taufik, M

AU - Waskita, Arya Adhyaksa

AU - Apriyanti, Diah Harnoni

PY - 2019/5/23

Y1 - 2019/5/23

N2 - Comparison to RGB, HSV, LAB and YCbCr color model in a flower recognition has been conducted for 12 species from two families. Flower images were obtained from the ImageCLEF 2017 dataset. After segmenting the flower from its background using the k-means clustering method, the statistical parameters are extracted from each color model. These parameters including the minimum and maximum pixel value, and also mean and standard deviation. Then, the classifying result of SVM shows that the HSV color model performs the worst amongst the other investigated color model, which is around 30% using different kernel. While the other color model has the accuracy around 70% using linear and polynomial kernel. This result in line with the surface plot of their statistical characteristics.

AB - Comparison to RGB, HSV, LAB and YCbCr color model in a flower recognition has been conducted for 12 species from two families. Flower images were obtained from the ImageCLEF 2017 dataset. After segmenting the flower from its background using the k-means clustering method, the statistical parameters are extracted from each color model. These parameters including the minimum and maximum pixel value, and also mean and standard deviation. Then, the classifying result of SVM shows that the HSV color model performs the worst amongst the other investigated color model, which is around 30% using different kernel. While the other color model has the accuracy around 70% using linear and polynomial kernel. This result in line with the surface plot of their statistical characteristics.

KW - image colour analysis

KW - feature extraction

KW - support vector machines

U2 - 10.1109/ICITISEE.2018.8721026

DO - 10.1109/ICITISEE.2018.8721026

M3 - Conference contribution

SN - 978-1-5386-7083-5

SP - 10

EP - 15

BT - 3rd International Conference on Information Technology, Information System and Electrical Engineering 2018

PB - IEEE

ER -