BIG-OH: BInarization of gradient orientation histograms
作者：
Highlights：
• BIG-OH, binary quantization of gradient orientation based descriptors, is proposed.
• Quantized SIFT descriptors reduce memory by 88% compared to classical SIFT.
• BIG-OH has performance comparable to SIFT and GLOH.
• BIG-OH has better performance than BRISK, CARD, BRIEF, and other descriptors.
• BIG-OH is effective for large scale applications such as copy detection.
摘要
•BIG-OH, binary quantization of gradient orientation based descriptors, is proposed.•Quantized SIFT descriptors reduce memory by 88% compared to classical SIFT.•BIG-OH has performance comparable to SIFT and GLOH.•BIG-OH has better performance than BRISK, CARD, BRIEF, and other descriptors.•BIG-OH is effective for large scale applications such as copy detection.
论文关键词：Gradient orientation histograms,SIFT,Gradient based keypoint descriptors,Keypoint descriptor quantization
论文评审过程：Received 20 May 2013, Revised 24 July 2014, Accepted 22 August 2014, Available online 30 August 2014.
论文官网地址：https://doi.org/10.1016/j.imavis.2014.08.006
