What do the white squares in the Kodak image represent?
yang991178
Those are pixels without a color filter at all, which allows the mixture of all color channels to pass through the microlen.
alstonchan64
This article sums up higher dynamic range and wider color gamut well. In this longer article, you can read about the tradeoffs between HDR standards (HDR10 vs. Dolby Vision).
Leon-Shao
Isn't it counter-intuitive that since humans are more sensitive to green portions, we actually should reduce the amount of green pixels in order to get a "better" color representation for human eyes?
hku8
how do the patterns vary when using CMYK versus RGB? Would it possibly be similar to the Sony RBG + E, as it adds an additional color to RGB?
briana-jin-zhang
@Leon-Shao, it looks like there are actually more green pixels in Bayer and Kodak's for that very reason.
selinafeng
If we used the bilinear interpolation demosaicking algorithm that we discussed in lecture on the Kodak pattern, how might the resulting image look different? Would the resolution and aliasing be worse than the other patterns?
alexkassil
[Delete]
alexkassil
There's a great data 100 project that showcases that humans are more sensitive to green channel in rgb. By adding the same amount of random noise to the red, blue, and green channel seperately, you could see the green image much more distorted
What do the white squares in the Kodak image represent?
Those are pixels without a color filter at all, which allows the mixture of all color channels to pass through the microlen.
This article sums up higher dynamic range and wider color gamut well. In this longer article, you can read about the tradeoffs between HDR standards (HDR10 vs. Dolby Vision).
Isn't it counter-intuitive that since humans are more sensitive to green portions, we actually should reduce the amount of green pixels in order to get a "better" color representation for human eyes?
how do the patterns vary when using CMYK versus RGB? Would it possibly be similar to the Sony RBG + E, as it adds an additional color to RGB?
@Leon-Shao, it looks like there are actually more green pixels in Bayer and Kodak's for that very reason.
If we used the bilinear interpolation demosaicking algorithm that we discussed in lecture on the Kodak pattern, how might the resulting image look different? Would the resolution and aliasing be worse than the other patterns?
[Delete]
There's a great data 100 project that showcases that humans are more sensitive to green channel in rgb. By adding the same amount of random noise to the red, blue, and green channel seperately, you could see the green image much more distorted