Object Recognize by Column Vector Based on Rearrange Chain-Code
This study deals with how to retrieve images using a rearranged chain-code based column vector. The chain-code expresses objects or boundaries of areas in straight chains with fixed direction and length, and encodes the final boundary, which is sensitive to rotated images. This study, therefore, works with chain-codes of eight directions in images and uses differences between values of chain-codes to obtain newly-rearranged chain-codes for image retrieval.
Y. An et al., "Object Recognize by Column Vector Based on Rearrange Chain-Code", Applied Mechanics and Materials, Vols. 20-23, pp. 768-773, 2010