Paper published in International Journal of Computer Vision Special Issue on Large Scale 3D Space

Scanning Depth of Route Panoramas Based on Stationary Blur (pdf)

Jiang Yu Zheng and Min Shi

Department of Computer Science

Indiana University Purdue University Indianapolis (IUPUI)

jzheng [at] cs.iupui.edu

A Java Demo to See Route Panoramas

 

 

 

 

 

 

 

 

 

After the image shows up, click the green words to start move. Also, click the yellow button to stop. Due to the network flow and the JAVA version (or maker) you are using, this traversing may take different actions. Sometimes, you may experience a (network) traffic jam.

Abstract

This work achieves an efficient acquisition of scenes and their depths along long streets. A camera is mounted on a vehicle moving along a straight or a mildly curved path and a sampling line properly set in the camera frame scans the 1D images over scenes continuously to form a 2D route panorama. This paper proposes a method to estimate the depth from the camera path by analyzing a phenomenon called stationary blur in the route panorama. This temporal blur is a perspective effect in parallel projection yielded from the sampling slit with a physical width. We analyze the behavior of the stationary blur with respect to the scene depth, vehicle path, and camera properties. Based on that, we develop an adaptive filter to evaluate the degree of the blur for depth estimation, which avoids error-prone feature matching or tracking in capturing complex street scenes and facilitates real time sensing. The method also uses much less data than the structure from motion approach so that it can extend the sensing area significantly. The resulting route panorama with depth information is useful for urban visualization, monitoring, navigation, and modeling.

Route Panorama and Measured Depth Map in the Following

(1) is a segment of route panorama. We can notice stationary blur at distant scenes.

(2) With differential operators, we can measure temporal contrasts in the route panorama with respect to the original contrasts in the image so that depth information can be figured out. This image shows the scene depths estimated by different filters. Here the depth values from three filters are displayed in R, G, B channels in order from small to large. The color shows which filter outputs the better results. Red: small filter, Yellow: small and median filters, Cyan: media and large filters, Blue: large filter. Intensity shows the depth value with bright values on close scenes and dark values on distant ones.

(3) The peaks of the spatial-temporal edges are located and their depths are obtained from the adaptive filtering.

(4) Rectified route panorama by removing waves caused by vehicle roll.

 

Filled Depth of Route Panorama along a Long Path (pdf)

 

> font>