Title :
Color saliency for object identification
Author :
Ozasa, Yuko ; Enami, Naoko ; Ariki, Yasuo
Author_Institution :
Grad. Sch. of Syst. Eng., Kobe Univ., Kobe, Japan
Abstract :
In an aging society, a service robot will come into our life. It is important for a robot to identify an object specified by human speech from several objects. Human may request an object for the robot by its name, and/or color name etc. Although there are some research about the method for the object identification based on its name, the object identification based on its color is not discussed enough. In this paper, we propose a method for the object identification by color name. The proposed method detects a saliency based on a color name, and the saliency is used for the identification.
Keywords :
image colour analysis; object recognition; robot vision; service robots; color name; color saliency; object identification; robot identification; service robot; Accuracy; Correlation; Histograms; Image color analysis; Object recognition; Robots; Training data;
Conference_Titel :
Frontiers of Computer Vision (FCV), 2015 21st Korea-Japan Joint Workshop on
Conference_Location :
Mokpo
DOI :
10.1109/FCV.2015.7103723