Improvement of Environment Recognition using Color Attribute Information

使用颜色属性信息改进环境识别

• 批准号: 10650266
• 负责人: KURATA Junichi
• 金额: $ 1.28万
• 依托单位: Kansai University
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for Scientific Research (C)
• 财政年份: 1998
• 资助国家: 日本
• 起止时间: 1998 至 1999
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/grant/KAKENHI-PROJECT-10650266/
• 关键词: color attribute; color image processing; color measurement; service robot; home cate technology; 色属性; 視覚センサ

In this study, the application of the perceptual information, which is based on the human attribute on color and informs quantitatively, has been described in order to make common information between the human and the machine and improve the information processing, especially color image processing.It seems to be difficult to recognize the similar color with different vividness and brightness by the combination of RGB information, even if that is very suitable information for the internal information of a visual equipment. Because, all of those are physically based on the amount of light with different wavelength, and they have same attribute from this point. On the other hand, the perceptual information "HSI", which deals with "Hue", "Saturation" and "Intensity", is very suitable to the color measurement tested by a visual inspection, but unsuitable for the internal information of a visual equipment. Because, these attribute have different characteristics each other. Therefore, the color measurement system, which has new information outputs by converting the RGB information into the perceptual information, has been developed in order to deal with color information easy. Since the proposed method would be expected to be equally used to measure the color more and more as the performance of imaging devises will be increasing, it has been applied to the some example in the measurement field.As the concrete example of environment recognizing sensor, the stereo range finder and the database of human sense on color were discussed. And also, the results of the test paper for urinalysis has been quantitatively measured by using the hue of color attribute, instead of the qualitative test by the grading according to several ranks. The experimental results of these proposals show the validity of proposed method using the perceptual information on color as one of the new measurement methodology.

为了使人与机器之间的信息共享，提高信息处理，特别是彩色图像的处理水平，本研究描述了基于人对颜色的属性并定量告知的感知信息的应用。即使RGB信息是非常适合视觉设备内部信息的信息，但通过RGB信息的组合似乎很难识别出生动度和亮度不同的相似颜色。因为，所有这些物理上都是基于不同波长的光的数量，从这一点上讲，它们具有相同的属性。另一方面，涉及“色相”、“饱和度”和“强度”的感知信息“HSI”非常适合视觉检测测试的颜色测量，但不适合视觉设备的内部信息。因为，这些属性彼此具有不同的特征。因此，为了方便地处理颜色信息，开发了将RGB信息转换为感知信息的新的信息输出的颜色测量系统。随着成像设备性能的不断提高，所提出的方法将被越来越多地应用于颜色的测量中，并已应用于测量领域的一些实例。作为环境识别传感器的具体实例，讨论了立体测距仪和人眼色彩感知数据库。用颜色属性的色相对尿检试纸的结果进行定量测量，而不是以往的按等级分级的定性测试。实验结果表明，将颜色感知信息作为一种新的测量方法是有效的。

项目成果

川崎安敦,藤原健吾,倉田純一,内山寛信: "自然言語による移動ロボットの誘導制御"日本機械学会ロボティクス・メカトロニクス講演会'99講演論文集. No.99-9. 2P2-23-010(1)-2P2-23-010(2) (1999)

Yasutoshi Kawasaki、Kengo Fujiwara、Junichi Kurata、Hironobu Uchiyama：“使用自然语言的移动机器人的引导和控制”日本机械工程师学会机器人和机电一体化会议99号会议记录。 1) -2P2-23-010(2) (1999)

Kawasaki Yasunobu, Kurata Junichi, Uchiyama Hironobu: "Guidance Control of a Service Robot"Preprint of JSME Kansai Conference. 14-14-14-14 (1999)

川崎康信、仓田淳一、内山博信：“服务机器人的引导控制”JSME关西会议预印本。

川崎、倉田、内山: "サービスロボツトの誘導制御" 日本機械学会関西支部第74期関西学生会卒業研究発表講演会論文集. (セッション14). 14 (1999)

川崎、仓田和内山：“服务机器人的引导和控制”日本机械工程师学会关西分会第 74 届关西学生会毕业研究报告讲座论文集（第 14 期）。

川崎安敦,倉田純一,内山寛信: "サービスロボットの誘導制御"日本機械学会関西学生会平成10年度卒業研究発表講演会前刷集. 14-14-14-14 (1999)

Yasutoshi Kawasaki，Junichi Kurata，Hironobu Uchiyama：“服务机器人的引导和控制”日本机械工程师学会关西学生会1998年毕业研究报告预印本集14-14-14-14（1999）。

常松美都子,山本修三,倉田純一,内山寛信: "知覚色情報に基づく簡易モニターシステムの開発"日本機械学会関西学生会平成12年度卒業研究発表講演会前刷集. 11-14-11-14 (2001)

Mitsuko Tsunematsu、Shuzo Yamamoto、Junichi Kurata、Hironobu Uchiyama：“基于感知颜色信息的简单监控系统的开发”日本机械工程学会关西学生会 2000 年毕业研究演示讲座预印本集 11-14-11-14。 ）