Image manipulation program built using Rust, wgpu and egui

All computation is performed in parallel on the GPU

The techniques implemented are:

• Gamma Correction
• Nearest Neighbor Interpolation
• Bilinear Interpolation
• Cross Correlation with Laplacian (by default)

The kernel used for cross correlation can be modified to use any values

New Images can be loaded using the Load button

Currently the implemented technique for to find the minimum and maximum for use in the normalization step is slow and inefficient. I tried implemented Parallel Reductions as described in GPU Gems Chapter 37 but was unable to get it to work.

Save functionality would be nice as well.

For a native binary:

cargo build --release
cargo run --release

For a native web version this can be compiled to WASM using wasm-pack

wasm-pack can be installed via cargo

wasm-pack build --target web