随笔分类 -  CV

SUSAN-FAST corner detectors
摘要：http://link.springer.com/article/10.1023/A:1007963824710#page-2 阅读全文

posted @ 2013-12-10 18:40 Aliceye 阅读(107) 评论(0) 推荐(0)

MSER-Maximally Extremal Regions –invariant contours
摘要：维基百科：http://zh.wikipedia.org/wiki/%E6%9C%80%E5%A4%A7%E7%A8%B3%E5%AE%9A%E6%9E%81%E5%80%BC%E5%8C%BA%E5%9F%9F 阅读全文

posted @ 2013-12-10 18:34 Aliceye 阅读(192) 评论(0) 推荐(0)

surf
摘要：http://blog.csdn.net/anson2004110/article/details/12951841 阅读全文

posted @ 2013-12-10 18:20 Aliceye 阅读(107) 评论(0) 推荐(0)

Harris
摘要：http://blog.csdn.net/linnol/article/details/6862005 阅读全文

posted @ 2013-12-10 18:17 Aliceye 阅读(88) 评论(0) 推荐(0)

SIFT
摘要：博客:http://www.dakaren.com/index.php/archives/639.htmhttp://blog.csdn.net/abcjennifer/article/details/7681718PCA-SIFThttp://blog.csdn.net/abcjennifer/article/details/7681718 阅读全文

posted @ 2013-12-10 11:24 Aliceye 阅读(118) 评论(0) 推荐(0)

Multiple View Geometry
摘要：On-line tutorial:在线教程 http://www.cs.unc.edu/~marc/tutorial.pdf http://www.cs.unc.edu/~marc/tutorial/ 阅读全文

posted @ 2013-12-10 09:48 Aliceye 阅读(150) 评论(0) 推荐(0)

lec6
摘要：See notes: Shu, Roth COMP 4900 slides • comp4900_lect5_filter1.pdf• comp4900_ lect6_filter2.pdf• comp4900_ lect7_feature1.pdf• comp4900_ lect8_feature2.pdfRANSACKLT tracker-------------------------------------------------------------------The Comparison and Application of Corner Detection Algorithms 阅读全文

posted @ 2013-12-08 19:00 Aliceye 阅读(193) 评论(0) 推荐(0)

lec5
摘要：还没整理....Image formats and noise•See notes: Shu, Roth COMP 4900 slides lect6_filter1.pdfGaussian smoothing•See notes: Shu, Roth COMP 4900 slides lect6_filter2.pdf***********edge***********See notes on edge detection:------------------------------------------------------------------------------------- 阅读全文

posted @ 2013-12-07 16:08 Aliceye 阅读(292) 评论(0) 推荐(0)

Finding Correspondences lec4
摘要：similarity measure********************SAD: Sum of Absolute DifferencesSAD=Σin|Xi-Yi| (similar lighting)some distinct points work well others don’t such as: edges, areas of uniformity (some lighting changes and slight (2 degrees)rotation)SAD adds up constant lighting changes.If two images are identi. 阅读全文

posted @ 2013-12-07 15:05 Aliceye 阅读(146) 评论(0) 推荐(0)

KLT
摘要：*********************KLT*********************http://www.cnblogs.com/moondark/archive/2012/05/12/2497391.html　先说KLT算法的几个前提假设： 1）亮度恒定 2）时间连续或者是运动是“小运动” 3）空间一致，临近点有相似运动，保持相邻Kanade-Lucas两人在上世纪80年代发表的paper：An Iterative Image Registration Technique with an Application to Stereo Vision这里讲的是一种图像点定位的方... 阅读全文

posted @ 2013-12-07 14:54 Aliceye 阅读(409) 评论(0) 推荐(0)

TLD
摘要：PN-learning:Bootstrapping Binary Classifiers by Structural Constraints, Zdenek Kalal,CVPR2010PN学习可以定义为一下的过程：（1）准备一个数量较少的训练样本集合和一个数量很大的测试样本集合。（2）利用训练样本训练一个初始分类器。同时，利用训练样本对(先验)约束条件进行相应的调整。（3）利用分类器对测试样本赋予标签，并找出分类器赋予的标签同约束条件相矛盾的那些样本；（4）将上述相矛盾的样本重新赋予标签，将其加入训练样本，重新训练分类器；反复迭代上述过程，直到满足某个约束条件Exploiting unlab 阅读全文

posted @ 2013-12-07 11:58 Aliceye 阅读(390) 评论(0) 推荐(0)

导航