Analysis and simulation of multipath separation based on improved penalty function method in time-of-flight cameras
Abstract
Time-of-flight cameras use the flight time of light to obtain accurate depth information, but its accuracy is affected by multipath interference. In the penalty function method adopted in this study, the best initial iteration point is obtained by cross-correlation of a pseudo-random binary sequence and its displacement sequence, and a continuous and non-segmented exponential penalty (obstacle) function is constructed, which can be used for sparse decomposition of high-dimensional vectors. These two improvements make the penalty function method applicable to sparse decomposition in the multipath separation process. Moreover, under the same measurement times and operation speeds, the improved penalty function method is nearly twice as accurate as the orthogonal matching pursuit algorithm in multipath separation.
- Publication:
-
Sixth Symposium on Novel Optoelectronic Detection Technology and Applications
- Pub Date:
- April 2020
- DOI:
- 10.1117/12.2560132
- Bibcode:
- 2020SPIE11455E..19J