PCA-SIFT: a more distinctive representation for local image descriptors

TLDR

Stable local feature detection and representation underpins many image registration and object recognition algorithms, and prior evaluations have shown SIFT to be the most robust to common image deformations. The study aims to improve the local image descriptor employed by SIFT. The authors replace SIFT’s weighted histograms with PCA on normalized gradient patches to encode salient image gradients. Experiments show that PCA-based descriptors are more distinctive, robust, and compact than SIFT, and that they improve accuracy and speed in image retrieval.

Abstract

Stable local feature detection and representation is a fundamental component of many image registration and object recognition algorithms. Mikolajczyk and Schmid (June 2003) recently evaluated a variety of approaches and identified the SIFT [D. G. Lowe, 1999] algorithm as being the most resistant to common image deformations. This paper examines (and improves upon) the local image descriptor used by SIFT. Like SIFT, our descriptors encode the salient aspects of the image gradient in the feature point's neighborhood; however, instead of using SIFT's smoothed weighted histograms, we apply principal components analysis (PCA) to the normalized gradient patch. Our experiments demonstrate that the PCA-based local descriptors are more distinctive, more robust to image deformations, and more compact than the standard SIFT representation. We also present results showing that using these descriptors in an image retrieval application results in increased accuracy and faster matching.