Robust regression and outlier detection
Robust regression and outlier detection
Two-dimensional signal and image processing
Two-dimensional signal and image processing
Hi-index | 0.00 |
Noise reduction (NR) in image sequences can improve both image quality and performance of subsequent video coding [1]-[4]. This paper proposes a new noise reduction system, MMOT-NR. It consists of a multi-frame block-based motion estimator which employs the least trimmed square (LTS) algorithm [5], a pixel-based trajectory correction unit, an outlier rejection unit, an NR unit, and an NR breakdown detector. Simulation results from HDTV color images are given to demonstrate the effectiveness of this system.