Volume 11, Number 3
Chiman Kwan and Jude Larkin, Applied Research, USA
The two mast cameras (Mastcam) act as eyes of the NASA’s Mars rover Curiosity. They can work independently or together for near and long range (up to 1 km) rover guidance and rock sample selection. Currently, the Mastcams are using Bayer color filter array (CFA), also known as CFA 1.0, in generating the RGB images. Under normal lighting conditions, CFA 1.0 is sufficient. However, since Mastcam may need to collect images under different lighting conditions such as early morning and sunset hours or sand storm periods, the lighting conditions in those scenarios will be unfavorable. It will be good to investigate a CFA that is robust to various lighting conditions. In the past, we have compared CFA 1.0 and CFA 2.0 for normal and low lighting images. Recently, a new CFA known as CFA 3.0 has been proposed by our team. CFA 3.0 has 75% white pixels, which are believed to be able to enhance the sensitivity of cameras. In this paper, we will first review some past demosaicing results for Mastcams. We will then investigate the performance of CFA 3.0 for Mastcam images in normal lighting conditions. Experiments using actual Mastcam images show that the demosaicing image quality using CFA 3.0 is satisfactory based on objective and subjective evaluations.
Mastcam, debayering, demosaicing, color filter array (CFA), RGBW pattern, Bayer pattern, CFA 1.0, CFA2.0, CFA3.0, pansharpening, deep learning