Gamma correction

[osman-turan]

At planned features, the gamma correction is listed. At the same time, the codebase has a commented-out piece of code for gamma correction. See:

photon/crate/src/colour_spaces.rs

Lines 10 to 50 in 475562b

	/// Apply gamma correction.
	// #[wasm_bindgen]
	// pub fn gamma_correction(mut photon_image: &mut PhotonImage, red: f32, green: f32, blue: f32) {
	// let img = helpers::dyn_image_from_raw(photon_image);
	// let (width, height) = img.dimensions();
	// let mut img = img.to_rgba8();
	// // Initialize gamma arrays
	// let mut gammaR: Vec<u8> = vec![];
	// let mut gammaG: Vec<u8> = vec![];
	// let mut gammaB: Vec<u8> = vec![];
	// let MAX_VALUE_INT = 255;
	// let MAX_VALUE_FLT = 255.0;
	// let REVERSE = 1.0;
	// // Set values within gamma arrays
	// for i in 0..256 {
	// gammaR[i] = min(MAX_VALUE_INT, ((MAX_VALUE_FLT * ((i as f32 / MAX_VALUE_FLT) as u32).powf(REVERSE / red) + 0.5 ) as u8));
	// gammaG[i] = min(MAX_VALUE_INT, ((MAX_VALUE_FLT * ((i as f32 / MAX_VALUE_FLT) as u32).powf(REVERSE / green) + 0.5 ) as u8);
	// gammaB[i] = min(MAX_VALUE_INT, ((MAX_VALUE_FLT * ((i as f32 / MAX_VALUE_FLT) as u32).powf(REVERSE / blue) + 0.5 ) as u8);
	// }
	// for x in 0..width {
	// for y in 0..height {
	// let px_data = img.get_pixel(x, y).data;
	// let r_val = px_data[0];
	// let g_val = px_data[1];
	// let b_val = px_data[2];
	// px_data[0] = gammaR[r_val as usize];
	// px_data[1] = gammaG[g_val as usize];
	// px_data[2] = gammaB[b_val as usize];
	// img.put_pixel(x, y, px);
	// }
	// }
	// photon_image.raw_pixels = img.to_vec();
	// }

I haven't checked the correctness of the commented-out function. Why is it commented-out? What's the expectation and plan for the gamma correction? I can work on this feature if you like.

[silvia-odwyer]

@osman-turan Hi Osman,
Thanks for opening this issue and for your interest in working on a gamma correction function!

The current gamma correction code uses features in the image library that are now deprecated, namely accessing the data field of each pixel, hence why the code is commented out. As for the correctness of the algorithm, I haven't been able to test it due to the deprecated code, so it might be better to start a new function and work on the implementation there.

But that said, if you'd like to work on your own function or make use of the logic in this, it's completely up to you! 😄 Thanks again for the help, it's greatly appreciated!

[osman-turan]

Thanks for the details! Due to rounding, there might be some edge cases at my implementation. If you encounter such inputs, please let me know so that I can work on a fix.