Feature Extraction of Real-Time Image Using SIFT Algorithm

Abstract

This paper deals with image processing and feature extraction. Feature extraction plays a vital role in the field of image processing. There exist different image pre-processing approaches for feature extraction such as binarization, thresholding, resizing, normalisation so on...Then after these techniques are applied to obtain high clarity images. In Feature extraction object recognition and stereo matching are at the base of many computer vision problems. The descriptor generator module is changed for increasing the performance of algorithm. SIFT algorithm consist of two modules such as key point detection module and descriptor generation module. When compared to recent solution, the descriptor generation module speed is fifteen times faster and the time for feature extraction is also reduced.

Performance comparison of ORB, SURF and SIFT using Intracranial Haemorrhage CTScan Brain images

“…The Algorithm must be resistant to noise and must be able to identify features quickly. The calculation speed of the algorithm must be very slow for use in modern equipment and systems [12]. Feature point comparison based on BRISK and ORB algorithms and algorithm improvement and feature point extraction experiments combining the advantages of the two are carried out.…”

Implementasi Advanced Driver Assistance System Menggunakan Metode Surf Dalam Mengenali Rambu-Rambu Lalu Lintas Di Indonesia

“…Beberapa algoritma ekstraksi fitur selain SURF adalah metode Oriented Fast and Rotated BRIEF (ORB) yang dapat melakukan pengenalan objek dan rekotruksi 3D [6] serta metode Scale Invariant feature Transform (SIFT) yang dapat melakukan pendeteksian gambar multiskala [7]. Pendeteksian suatu objek diperlukan titik korespondensi antara dua gambar dari setiap scene atau objek yang sama.…”