Realistic modeling of a 3D environment has grown in popularity due to the increasing realm of practical applications. Whether for practical navigation purposes, entertainment value, or architectural standardization, the ability to determine the dimensions of a room is becoming more and more important. One of the trickier, but critical, features within any multistory environment is the staircase. Staircases are difficult to model because of their uneven surface and various depth aspects. Coupling this need is a variety of ways to reach this goal. Unfortunately, many such methods rely upon specialized sensory equipment, multiple calibrated cameras, or other such impractical setups. Here, we propose a simpler approach.
This paper outlines a method for extracting the slope dimensions of a staircase using a single monocular image. By relying on only a single image, we negate the need for extraneous accessories and glean as much information from common pictures. We do not hope to achieve the high level of accuracy seen from laser scanning methods but seek to produce a viable result that can both be helpful for current applications and serve as a building block that contributes to later development.
When constructing our pipeline, we take into account several options. Each step can be achieved with different techniques which we evaluate and compare on either a qualitative or quantitative level. This leads to our final result which can accurately determine the slope of a staircase with an error rate of 31.1%. With a small amount of previous knowledge or preprocessing, this drops down to an average of 18.7% Overall, we deem this an acceptable and optimal result given the limited information and processing resources which the program was allowed to utilize.
Identifer | oai:union.ndltd.org:CALPOLY/oai:digitalcommons.calpoly.edu:theses-2563 |
Date | 01 June 2015 |
Creators | Clarke, Nicholas Joseph |
Publisher | DigitalCommons@CalPoly |
Source Sets | California Polytechnic State University |
Detected Language | English |
Type | text |
Format | application/pdf |
Source | Master's Theses |
Page generated in 0.0135 seconds