Assessing Carpet Appearance Retention by Image Analysis

Abstract

Emphasis on carpet appearance retention in recent research has created an interest in developing reliable and efficient methods for evaluating carpet appearance. A number of image analysis techniques have been applied for this purpose. The current study attempts to expand application of image analysis to assessing appearance change in tufted carpets by means of tuft characterization, since tuft characteristics such as ge ometry, orientation, and spatial arrangement are key factors in visual assessment and therefore in image analysis of carpet appearance. Beginning with a brief review of image analysis methods previously applied, this paper focuses on the uses of three techniques—shape, orientation, and spatial pattern analyses—for tuft characterization. Digital image processing techniques necessary for enhancing carpet images are also introduced. The experimental materials consist of two sets of wool carpets and two sets of nylon carpets of different wear levels.