Title :
Research on Image Based on Improved SURF Feature Matching
Author :
Yanling Liu ; Sihang Ma
Author_Institution :
Tianjin Key Lab. of the Design & Intell. Control of the Adv. Mechatronical Syst., Tianjin Univ. of Technol., Tianjin, China
Abstract :
Image matching is widely used in machine vision, computer technology, industry and agriculture and other fields. Matching the characteristics of traditional methods such as SIFT has advantages of compressed information quantity, high precision, but there are also some deficiencies such as the large of calculation, long time, the object to be measured of position should be required accurately and etc. This paper presents image matching method based on improved SURF feature, and the analysis and comparison are made for the three kinds of image matching methods. The experiment proved: Sometimes it can also be A single SIFT or SURF, The improved SURF more can meet the requirements of parameter estimation accuracy and real-time. It can carry out accurate judgment, recognition for the digital image. It has certain theoretical value and practical value.
Keywords :
feature extraction; image matching; transforms; SIFT; compressed information quantity; computer technology; digital image recognition; image matching; improved SURF feature matching; machine vision; parameter estimation; scale-invariant feature transform; speeded up robust features; Accuracy; Algorithm design and analysis; Feature extraction; Histograms; Image matching; Signal processing algorithms; Vectors; SIFT; SURF; image matching; machine vision;
Conference_Titel :
Computational Intelligence and Design (ISCID), 2014 Seventh International Symposium on
Print_ISBN :
978-1-4799-7004-9
DOI :
10.1109/ISCID.2014.133
