TY - GEN
T1 - BIM
T2 - 2009 IEEE International Conference on Image Processing, ICIP 2009
AU - Kim, Hung Sik
AU - Chang, Hau Wen
AU - Liu, Haibin
AU - Lee, Jeongkyu
AU - Lee, Dongwon
PY - 2009
Y1 - 2009
N2 - Matching two images with similar contents is one of the most fundamental tasks in image processing. Due to its importance, in recent years, many novel techniques have been proposed with great successes. Toward this effort, in this paper, we propose a radically different idea by bridging two seemingly unrelated fields - Image Processing and Biology - i.e., we propose to use the popular gene sequence alignment algorithm in Biology, BLAST, in determining the similarity between images. In this proposal, we map image features to a sequence of gene alphabets (e.g., A, C, G, and T in DNA, or 23 letters in protein) to utilize a wealth of advanced algorithms and tools in BLAST. Under the new idea, in particular, we study various image features and gene sequence generation methods that impact the accuracy and performance in matching similar images. Our proposal, termed as BLASTed Image Matching (BIM), is empirically validated using real data sets. Our work can be viewed as the "first" step toward bridging Image Processing and Biology fields in the application of the well-studied image matching problem.
AB - Matching two images with similar contents is one of the most fundamental tasks in image processing. Due to its importance, in recent years, many novel techniques have been proposed with great successes. Toward this effort, in this paper, we propose a radically different idea by bridging two seemingly unrelated fields - Image Processing and Biology - i.e., we propose to use the popular gene sequence alignment algorithm in Biology, BLAST, in determining the similarity between images. In this proposal, we map image features to a sequence of gene alphabets (e.g., A, C, G, and T in DNA, or 23 letters in protein) to utilize a wealth of advanced algorithms and tools in BLAST. Under the new idea, in particular, we study various image features and gene sequence generation methods that impact the accuracy and performance in matching similar images. Our proposal, termed as BLASTed Image Matching (BIM), is empirically validated using real data sets. Our work can be viewed as the "first" step toward bridging Image Processing and Biology fields in the application of the well-studied image matching problem.
UR - http://www.scopus.com/inward/record.url?scp=77951966477&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=77951966477&partnerID=8YFLogxK
U2 - 10.1109/ICIP.2009.5414214
DO - 10.1109/ICIP.2009.5414214
M3 - Conference contribution
AN - SCOPUS:77951966477
SN - 9781424456543
T3 - Proceedings - International Conference on Image Processing, ICIP
SP - 205
EP - 208
BT - 2009 IEEE International Conference on Image Processing, ICIP 2009 - Proceedings
PB - IEEE Computer Society
Y2 - 7 November 2009 through 10 November 2009
ER -