Instrumental Techniques to Quantify Textural and Appearance Changes in Carpet

Abstract

Our research emphasis has been on identifying several key features of carpet quality combining both objective (instrumental) and subjective (human perception, HPA) approaches. Previously, HPA was used to determine a set of user terms and descriptions for assessing carpet quality features such as tuft geometry, luster, soiling, and overall appearance. The work reported here expands on the development of objective or instrumental methods for quantifying user perception of carpet appearance. A cus tomized, personal computer-based image analysis system, using an AT&T image cap ture board, has been developed. This new endeavor permits moderate resolution video capture of color images, with digital storage, retrieval, measurement, analysis, and enhancement. Special programming and illumination specifically designed for carpet surface analyses as well as carpet fiber-yarn properties are underway. The work reported here includes measurement of yam bundle or tuft size distribution through a regime of image enhancement, binary conversion, edge detection, and area measurement. Data are presented on various worn and unworn polyester Saxony tufted carpets using these techniques. Major new findings demonstrate the importance of color (RGB) imaging or colorimetric image analysis, in addition to gray level image processing, for determining appearance changes in carpet.