Changeset 702 for branches/2010-image-rec/docs/report.tex
- Timestamp:
- 02/22/10 20:44:33 (2 years ago)
- Files:
-
- 1 modified
-
branches/2010-image-rec/docs/report.tex (modified) (2 diffs)
Legend:
- Unmodified
- Added
- Removed
-
branches/2010-image-rec/docs/report.tex
r701 r702 82 82 Our process to get from camera images to information useful for navigation 83 83 requires several steps of processing and reprocessing. First, the image is 84 converted to the HSV color space. We then use a color histogram classifier,84 converted to the HSV color space. We then use a color histogram classifier, 85 85 based on an assumed safe area of grass, to classify pixels in the image as 86 86 grass and not grass. Then the texture data from application of a Sobel filter … … 155 155 156 156 \section{Future work} 157 158 \begin{figure}[htb] 159 \centering 160 \includegraphics[width=0.4\textwidth]{figures/hough-lines-failure-mask.png} 161 \includegraphics[width=0.4\textwidth]{figures/hough-lines-failure.pdf} 162 \caption{An illustration of the failure modes of the present line detector. At left, the line mask has picked up on highlights on the barrel, producing false lines to be identified at right.} 163 \label{fig:hough-failure} 164 \end{figure} 165 157 166 The next step with our algorithm is to make it work on cases where it fails, 158 167 those with lines appearing within our safe zone. It is possible that this

