A Comparative Analysis of Depth-Discontinuity and Mixed-Pixel Detection Algorithms

摘要: Laser scanner measurements are corrupted by noise and artifacts that can undermine the performance of registration, segmentation, surface reconstruction, recognition, other algorithms operating on data. While much research has addressed laser models, comparatively little is known about artifacts, such as mixed pixel effect, color-dependent range biases, specular reflection effects. This paper focuses effect related challenge detecting depth discontinuities in 3D a number have been proposed for pixels discontinuities, there no consensus how well perform or which algorithm performs best. presents comparative analysis five mixed-pixel/discontinuity detection real data sets. We find an based normal angle best overall performance, but exceptionally well. Factors influencing also discussed.