SUMMARY
IntroductionIn computervision it is expected that the capability of recognizing an object is increased significantly by additional information of material types to conventional geometrical information such as the position, shape, and posture of the object. To materialize this idea, this paper proposes a simple color-reflection model which classifies colored opaque objects seen under white light according to their gloss into three categories: "metallic;" "matte nonmetallic;" and "glossy nonmetallic." This model is based on the dichromatic reflection model proposed byShafer but with an extended capability to include metallic objects. The model can recognize not only apparent material but can also detect interreflection of the object. The effectiveness of the method was confirmed successfully by applying it to actual scenes containing metallic and nonmetallic objects.Key words: Color-reflection model; metallic objects; nonmetallic objects; computervision; reflection characteristics; dichromatic reflection model.A process to recognize an object in computervision consists of the extraction of features in its image and comparing them with their models. In a conventional way, some geometrical features, such as the position, shape, and posture of an object, are used for the recognition of the object. However, there are some cases where an object is not represented clearly by the shape alone or it is difficult to make a model of an object such as a flexible object. To add the reflection characteristics of the surface of materials of an object to conventionally used features is one possible way to solve this problem. * A method of measuring the distribution of the lightabsorption rate on the surface of an object is most fundamental to discriminate its material using the reflection characteristics. However, this is too complex for computemision since it is necessary to measure spectra in the full visual light range. Therefore, obtaining material information using an RGB color image, which is still related t o the spectrum information, has attracted the attention of researchers recently.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.