Vision-based 3-D trajectory tracking for unknown environments

An Ego-Motion Detection System Employing Directional-Edge-Based Motion Field Representations

“…One of the major problems of interpreting ego-motion is that the camera rotation and translation can induce similar motion patterns [12], [36]- [38]. To deal with this ambiguity problem, additional constraints are usually required for motion models so as to disambiguate a true global motion direction [19], [39], [40].…”

“…Typically, there are two stages in ego-motion estimation: the extraction of motion information from a scene and the interpretation of the accumulated data. The first stage, the motion extraction stage, is usually done with one of either two approaches: feature correspondence method [7]- [12], or optical flow field method [13]- [19]. In the former, ego-motion is detected by extracting some clearly distinguishable features from a scene and tracking each of them from frame to frame.…”

A Visual Navigation Strategy Based on Inverse Perspective Transformation

“…Nister et al estimated the motion of mobile robots tracking nondistinctive Harris corners. Saeedi et al (Saeedi et al, 2006) presented a stereo vision-based 3-D trajectory tracker for localization of mobile robots in unexplored environments. This work proposed a new corner detector to extract image features.…”

Onboard Video Stabilization for Unmanned Air Vehicles

“…The first step of Binary Corner Detection is to smooth the image [35]. This helps reduce image noise [32].…”

“…This helps reduce image noise [32]. A Gaussian smoothing is applied with a σ value of 0.8 [35]. This value is strategic because it allows a filter to be approximated …”