TIME COST COMPARISON AT DIFFERENT
PHASES Star Student Project OF THE DISTRIBUTED/NON
DISTRIBUTED SEARCH SYSTEMS (MS). COMMUNICATION TRANSFER WORDS TO
DISTRIBUTED SERVERS Star Student Project COMMUNICATION II:
MERGE THE MULTIPLE RANKING LISTS FROM DISTRIBUTED SERVERS . MAXIMUM
SEARCH MAXIMUM SEARCH COST Star Student Project AMONG
DISTRIBUTED SERVERS. SIFT EXTRACTION IS PARALLELED BY DIVIDING THE
QUERY IMAGE INTO OVERLAPPING RECTANGLES, AS SHOWN IN Star Student
Project . DUPLICATE DETECTED LOCAL FEATURE ARE FILTERED OUT BY
COORDINATE DE-DUPLICATIONS, AS SHOWN IN Star Student Project .
NOTE THAT THE COMPLEXITY OF SIFT EXTRACTION IS INDEPENDENT OF DATASET
SCALE Star Student Project Distribute the process of local
feature extraction by partitioning a query image Star Student
Project into multiple overlapping rectangles followed by
individual rectangles based feature extraction. Star Student
Project servers. .
