Correction and interpolation of depth maps from structured light infrared sensors
作者：
Highlights：
• The paper presents a denoising strategy for infrared structured light depth sensors.
• Errors are evaluated by measuring the edge mismatch between color and depth images.
• Evaluation datasets are proposed with ground-truth point clouds from a laser scanner.
• Multiple acquisitions were included for different noise and illumination conditions.
• Experimental results prove that the solution performs well and in real time.
摘要
Highlights•The paper presents a denoising strategy for infrared structured light depth sensors.•Errors are evaluated by measuring the edge mismatch between color and depth images.•Evaluation datasets are proposed with ground-truth point clouds from a laser scanner.•Multiple acquisitions were included for different noise and illumination conditions.•Experimental results prove that the solution performs well and in real time.
论文关键词：DIBR denoising,MS Kinect,Structured light camera,Infrared sensor,3D acquisition
论文评审过程：Received 8 May 2015, Revised 25 November 2015, Accepted 30 November 2015, Available online 11 December 2015, Version of Record 22 December 2015.
论文官网地址：https://doi.org/10.1016/j.image.2015.11.008
