Cheap Upscaling Triangulation

Cheap Upscaling Triangulation (CUT) is a family of single-image upscaling algorithms for retro games designed to be:

• Versatile: can upscale from and to any image resolution and are applicable to all the 2D and 3D consoles that Lemuroid supports
• Efficient: keep the number of samples and calculations as low as possible to minimize battery consumption

In order to achieve this, we need to CUT some corners... Literally!

Algorithms

The family is composed of two algorithms CUT1 and CUT2, which share the same basic steps:

• Edge Detection
• Triangulation / Pattern Recognition
• Interpolation

These are implemented differently in the two algorithms, leading to different quality and performance levels:

• CUT1: Uses a 2x2 pixel window and can approximate edges of 45°
• CUT2: Uses a 4x4 pixel window and can approximate edges of 30°, 45° and 60°

Configuration

The look of both versions can be customized with a set of parameters:

#define USE_DYNAMIC_BLEND 1          // Dynamically blend color with respect to contrast
#define BLEND_MIN_CONTRAST_EDGE 0.0  // Minimum contrast level at which sharpness starts increasing [0, 1]
#define BLEND_MAX_CONTRAST_EDGE 1.0  // Maximum contrast level at which sharpness stops increasing [0, 1]
#define BLEND_MIN_SHARPNESS 0.0      // Minimum sharpness level [0, 1]
#define BLEND_MAX_SHARPNESS 1.0      // Maximum sharpness level [0, 1]
#define STATIC_BLEND_SHARPNESS 0.5   // Sharpness level used when dynamic blending is disabled [0, 1]
#define EDGE_USE_FAST_LUMA 0         // Use quick luma approximation in edge detection
#define EDGE_MIN_VALUE 0.05          // Minimum luma difference used in edge detection [0, 1]
#define EDGE_MIN_CONTRAST 1.20       // Minimum contrast ratio used in edge detection [1, ∞]
#define LUMA_ADJUST_GAMMA 0          // Correct gamma to better approximate luma human perception

Results

Here you can find some results. Each image is split in two, with the left side unprocessed and the right side with CUT2 applied.

You can also see some examples using CUT1.

Performances

There aren't yet extensive performance tests, but I tried measuring GPU load on my device, a Galaxy S21 FE with Snapdragon 888 playing Final Fantasy VI Advance.

Filter	GPU Utilization	Resolution
Bilinear (Lemuroid)	~1.5%	160p
HQx2 (Retroarch)	~2.5%	320p
CUT1 (Lemuroid)	~3.5%	1080p
HQx4 (Retroarch)	~4.5%	640p
CUT2 (Lemuroid)	~6.5%	1080p
xbrz-freescale-multipass (Retroarch)	~15.0%	1080p

References

• D. Su and P. Willis, "Image interpolation by pixel level data-dependent triangulation", Computer Graphics Forum, pp. 23, 2004.
• A. Reshetov, "Morphological antialiasing", Proceedings of the Conference on High Performance Graphics 2009, pp. 109-116, 2009.