Comparison of image matching techniques
International Journal of Latest Trends in Engineering and Technology · 2016
In today's times, cameras have become a major role player and can be seen everywhere, from the smart phone in our pocket to the surveillance cameras in our campus to the microscopic cameras used in medical sciences and so on.The field of computer vision has seen a meteoric rise in the recent past, with the development of a wide variety of techniques to accomplish certain tasks.These tasks include motion analysis, scene reconstruction, image restoration and image matching [1][2][3] .In this study, we have focused on various image matching techniques and algorithms.We have compared their performances, eventually suggesting the best technique out of all the considered techniques.It may happen that some of these algorithms/techniques work better with certain data sets, while others aren't as effective in analysing the same data sets.Hence, certain algorithms prove to be useful for a specific application while others have different usage.As mentioned in the text above, computer vision algorithms are widely used to recognize, manipulate and extract details from image data.These processes are conducted with the help of various algorithms and techniques.Each algorithm has its unique way of identifying and governing the data that is to be modified.Every algorithm is unique from the other one and efficiency criteria differ in each case, even though the aim of the algorithm is the same i.e. image matching.Image matching is a sub domain of computer vision, which focuses on finding a similarity or multiple