View-invariant object recognition using homography constraints

Abstract

Change in viewpoint is one of the major factors for variation in object appearance across different images. Thus, view-invariant object recognition is a challenging and important image understanding task. In this paper, we propose a method that can match objects in images taken under different viewpoints. Unlike most methods in the literature, no restriction on camera orientations or internal camera parameters are imposed and no prior knowledge of 3D structure of the object is required. We prove that when two cameras take pictures of the same object from two different viewing angels, the relationship between every quadruple of points reduces to the special case of homography with two equal eigenvalues. Based on this property, we formulate the problem as an error function that indicates how likely two sets of 2D points are projections of the same set of 3D points under two different cameras. Comprehensive set of experiments were conducted to prove the robustness of the method to noise, and evaluate its performance on real-world applications, such as face and object recognition.

Motion-Compensated Temporal Filtering for Critically-Sampled Wavelet-Encoded Images

“…Motion Estimation/Compensation and video coding have wide range of applications in various areas of image/video processing, including restoration [7], [32]- [34], [47], [67], [94], [107]- [109], [111]- [117], [120], [121], [123], content/context analysis [1]- [3], [37]- [39], [54], [55], [95], [101], [138]- [141], [147], surveillance [11], [16], [17], [69]- [71], [74]- [77], [84], [132], [134], [135], [137], action recognition [10], [13]- [15], [18], [19], [36], [124]- [129], [133], [136], self-localization [78], [79],…”

A Closed-Form Model for Image-Based Distant Lighting

“…Image-based rendering (IBR) has been an active area of research in computational imaging and computational photography in the past two decades. It has led to many interesting non-traditional problems in image processing and computer vision, which in turn have benefited from traditional methods such as shape and scene description [1]- [3], [10], [11], [15], [16], [34], [36]- [38], [54], [56], [69]- [72], [75]- [78], [83], [85], [95], [101], [132], [134], [135], [137], [146], scene content modeling [79], [80], [84], [86]- [88], [139]- [143], super-resolution (in particular in 3D) [31]- [33], [46], [67], [68], [94], [104]- [106], [108]- [114], [117], [118], [120], video content modeling [9], [12]-…”

An Invariant Model of the Significance of Different Body Parts in Recognizing Different Actions

“…Often, multiple view methods are designed to tackle viewpoint invariant recognition [9], [12], [15], [16], [113]- [115], [117], although such methods may require calibration across views [8], [25], [56], [57], [64], [65], [71]- [74], image registration [7], [17]- [19], [21]- [24], [27], [28], [43]- [49], [95], [103], [104], or tracking across views [79]- [81], [107], [118]. There are also methods that rely on human-object interaction [87], [135], [136], which often require identifying image contents other than humans [2]- [4], [36]- [38], [40], [42], [76], [78], [85], [134]. Other preprocessing steps that may be needed include image restoration [29]- [31], [53]<...>…”