Texture Classification with Minimal Training Images

 

People: Alireza Tavakoli Targhi, Jan-Mark Geusebroek, Andrew Zisserman


The objective of this work is classifying texture from a single image under unknown lighting conditions. The current and successful approach to this task is to treat it as a statistical learning problem and learn a classifier from a set of training images, but this requires a sufficient number and variety of training images. We show that the number of training images required can be drastically reduced (to as few as three) by synthesizing additional training data using photometric stereo. We demonstrate the method on the PhoTex and ALOT texture databases. Despite the limitations of photometric stereo, the resulting classification performance surpasses the state of the art results.


Over view of Photometric Stereo
































Overview of experience














Sample materials from Amsterdam ALOT Database. In reading order, with ALOT class number between brackets: teawafers (9), brown bread (26), cotton (43), terry cloth (48), punched plastic (56); cork (57), cotton (60), ribbed cotton (64), sponge (176), and chamois (196).



A snapshot of ALOT Database:  Link of the Database Page


  1. Number of class: 250

  2. Number of image per class: 100

  3. 8 illumination , 3 orientation, 4 view point











A snapshot of KTH-TIP2 Database:  Link of the Database Page

  1. Number of class: 11

  2. Number of image per class: 108

  3. 4 illumination , 9 scale, 3 orientation





























Sample of 9 scale of KTH-TIPS2

































 

Related Paper

Texture Classification with Minimal Training Images.

Alireza Tavakoli Targhi, Jan-Mark Geusebroek, Andrew Zisserman.

The 19th International Conference on Pattern Recognition, Florida, USA, 2008 (ICPR 2008).

[Paper PDF][Presentation PPT]