Volume 15, Issue 6 p. 872-880
Paper

Analyzing amber color features and methods of estimating colors of amber debris

Hina Ishikawa

Hina Ishikawa

Non-member

Graduate School of Engineering Science, Akita University, 1-1 Tegata Gakuen-machi, Akita-shi, Akita, 010-8502 Japan

Search for more papers by this author
Yoichi Kageyama

Corresponding Author

Yoichi Kageyama

Member

Graduate School of Engineering Science, Akita University, 1-1 Tegata Gakuen-machi, Akita-shi, Akita, 010-8502 Japan

Correspondence to: Yoichi Kageyama. E-mail: kageyama@ie.akita-u.ac.jpSearch for more papers by this author
Chikako Ishizawa

Chikako Ishizawa

Member

Graduate School of Engineering Science, Akita University, 1-1 Tegata Gakuen-machi, Akita-shi, Akita, 010-8502 Japan

Search for more papers by this author
Tsuyoshi Takahashi

Tsuyoshi Takahashi

Member

Graduate School of Engineering Science, Akita University, 1-1 Tegata Gakuen-machi, Akita-shi, Akita, 010-8502 Japan

Search for more papers by this author
Makoto Nishida

Makoto Nishida

Non-member

The Open University of Japan, 1-1, Tegata Gakuen-machi, Akita-shi, Akita, 010-8502 Japan

Search for more papers by this author
Hisao Shinden

Hisao Shinden

Non-member

Kuji Kohaku Co., Ltd, 19-156-133, Kokuji-cho, Kuji-shi, Iwate, 028-0071 Japan

Search for more papers by this author
Toshio Takizawa

Toshio Takizawa

Non-member

Kuji Kohaku Co., Ltd, 19-156-133, Kokuji-cho, Kuji-shi, Iwate, 028-0071 Japan

Search for more papers by this author
Taro Koyama

Taro Koyama

Non-member

Kuji Kohaku Co., Ltd, 19-156-133, Kokuji-cho, Kuji-shi, Iwate, 028-0071 Japan

Search for more papers by this author
Makoto Kobushi

Makoto Kobushi

Non-member

Kuji Kohaku Co., Ltd, 19-156-133, Kokuji-cho, Kuji-shi, Iwate, 028-0071 Japan

Search for more papers by this author
First published: 30 March 2020

Abstract

In our previous study, we presented a method of estimating the color of amber using the red, green, and blue (RGB) color space and the hue, saturation, and value (HSV) color space. In particular, the method estimates the color of amber on a target image consisting of four colors (yellow, brown, black, and amber with an outer cover [AOC]). However, because of the existence of amber with a partially exposed outer cover (APEOC), the estimated accuracy of AOCs was 57.1%. In addition, the size of the amber used in the study was approximately 1–2 cm, considering improvements in versatility. It is also necessary to estimate the color of amber with a size of approximately 3–5 mm, which is even smaller. However, because the color of brown amber is similar to yellow amber, the estimation accuracy of amber having a size of less than 1 cm, using conventional methods, was 62.1%. Therefore, in this paper, we propose a method that switches the estimation method in terms of color features and amber of sizes ranging between 1 and 2 cm and those of 1 cm or less. Based on the experimental results of the proposed method, estimation accuracies of 94.1% for 185 ambers with a size of 1–2 cm and 96.0% for 125 ambers with a less than 1 cm were achieved. © 2020 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.

The full text of this article hosted at iucr.org is unavailable due to technical difficulties.