Motion Vector Estimation of a Stereovision Camera with Inertial Sensors

Abstract

The aim of the presented work was the development of a tracking algorithm for a stereoscopic camera setup equipped with an additional inertial sensor. The input of the algorithm consists of the image sequence, angular velocity and linear acceleration vectors measured by the inertial sensor. The main assumption of the project was fusion of data streams from both sources to obtain more accurate ego-motion estimation. An electronic module for recording the inertial sensor data was built. Inertial measurements allowed a coarse estimation of the image motion field that has reduced its search range by standard image-based methods. Continuous tracking of the camera motion has been achieved (including moments of image information loss). Results of the presented study are being implemented in a currently developed obstacle avoidance system for visually impaired pedestrians.

Automatic Calibration of Stereoscopic Cameras in an Electronic Travel Aid for the Blind

“…Fig. 4a) shows characteristic image points for which the movement was verified by the gyroscope (9). It can be seen that points belonging to nearby objects were discarded.…”

“…The measure of certainty of a scene point being distant is expressed as the number of frames its motion is compatible with the model (9). Seemingly, this time should be as long as possible, but an excessive increase of the threshold results in too many points having the opportunity to leave the field of view of the cameras or get occluded by scene objects.…”

“…The disadvantage is that the algorithm requires continuous translational motion of the cameras for proper selection of characteristic distant scene points. There are plans to integrate the algorithm with an egomotion tracking technique [9] that could assess the suitability of consecutive frames for model building depending on the detected camera movement.…”

“…• Searching for characteristic points in the left image using a Harris corner detector, • Determination of the corner movement between consecutive images, verified with angular measurements from an electronic MG1101B gyroscope [9], • Determination of disparity for the corner points,…”

Stars’ Centroid Determination Using PSF-Fitting Method

“…The centroid detection has to be preceded by filtering out the image distortions. Many different computation methods are adopted to reduce the image distortions [7][8][9][10]. Mathematically simplest methods are based on computing the centre of mass [11] and the weighted centre of mass [12][13][14] of a star, respectively.…”