Automatic Stain Detection on Fabrics Using Image Processing

Dry cleaners often find it time consuming and difficult in finding the invisible stains and fabric flaws. An automatic stain detector in fabrics has been proposed in this paper. The algorithm uses k-means clustering methods to identify the stains. The clustering approach for stain detection provides an apparatus that allows stain detection and analysis of flowed or damaged fabrics in a quick and efficient manner. Another objective of the proposal is to provide a detector that is easy to use, light weight, portable, convenient and durable. The algorithm is useful to personnel in the areas of textiles, dry cleaners, garment analysts, museum curators, manufacturers and retailers of clothing, textiles and upholstered furniture. Another objective of the algorithm is to provide a method for accurately identifying stains on various types of fabrics. Segmentation approaches in image processing can be used for this purpose. Clustering methods in segmentation is used in this algorithm to provide a faster and easier algorithm for detection of stains. The cluster centres are assigned in a random manner to further calculate the clusters and to classify the stains and the fabric.