Global and local features for accurate impression estimation of cloth fabric images

Consumers´ psychological feeling or impression is an important factor for product design. The impression estimation becomes an important issue. In this paper, we proposed a machine learning based impression estimation method for cloth fabric images. We use a semantic differential (SD) method to measure the user´s impression such as bright, warm while they viewing a cloth fabric image. We also extract both global and local features of cloth fabric images such as color and texture using computer vision techniques. Then we use support vector regression to model the mapping function between the impression and image features. The learned mapping function is used to estimate the impression of cloth fabric images.