More translucent and darker at the incisal edge<\/li>\n<\/ul>\n\n\n\nMost digital shade apps reduce the entire tooth to a single averaged RGB value. That approach throws away critical information.<\/p>\n\n\n\n
Shade Master does something different.<\/p>\n\n\n\n
It treats shade selection as a multi-region color comparison problem<\/strong> using perceptually accurate color spaces.<\/p>\n\n\n\nThe Core Idea Behind the Algorithm<\/h1>\n\n\n\n Instead of comparing raw RGB values, Shade Master:<\/p>\n\n\n\n
\nSplits the tooth into multiple regions<\/li>\n\n\n\n Extracts representative colors from each region<\/li>\n\n\n\n Converts them to CIELAB (a perceptually uniform color space)<\/li>\n\n\n\n Compares full tonal distributions instead of single color points<\/li>\n\n\n\n Selects the shade with the lowest average \u0394 (Delta E)<\/li>\n<\/ol>\n\n\n\nThis is not just picking a color.
How the Algorithm Works<\/h1>\n\n\n\n Here\u2019s a simplified breakdown of the process.<\/p>\n\n\n\n
<\/p>\n\n\n\n<\/p>\n\n\n\n
Step 1 \u2014 Divide the Tooth Into Regions<\/h2>\n\n\n\n The clinical tooth image is divided into structured regions (for example: cervical, middle, incisal).<\/p>\n\n\n\n
Each region is processed independently.<\/p>\n\n\n\n
Step 2 \u2014 Extract Representative Colors<\/h2>\n\n\n\n For each region:<\/p>\n\n\n\n
\nRGB pixels are collected<\/li>\n\n\n\n Converted to CIELAB<\/li>\n\n\n\n Sorted by lightness (L*)<\/li>\n\n\n\n Outliers are pruned<\/li>\n\n\n\n The median brightness color is selected<\/li>\n<\/ul>\n\n\n\nWhy CIELAB?<\/p>\n\n\n\n
Because CIELAB models human color perception more accurately than RGB<\/strong>.Step 3 \u2014 Build a Unified Tooth Palette<\/h2>\n\n\n\n All cleaned regional palettes are:<\/p>\n\n\n\n
\nMerged<\/li>\n\n\n\n Re-sorted by brightness<\/li>\n\n\n\n Evenly sampled to match the size of shade region palettes<\/li>\n<\/ul>\n\n\n\nThis creates a structured tonal fingerprint<\/strong> of the tooth.<\/p>\n\n\n\nNot just one color \u2014 but a gradient signature.<\/p>\n\n\n\n
Step 4 \u2014 Process Each Shade Tab the Same Way<\/h2>\n\n\n\n Each candidate shade (A, B, C, D, etc.) goes through the same pipeline:<\/p>\n\n\n\n
\nRegion segmentation<\/li>\n\n\n\n RGB \u2192 CIELAB conversion<\/li>\n\n\n\n Brightness sorting<\/li>\n\n\n\n Outlier pruning<\/li>\n<\/ul>\n\n\n\nThe shade is now represented as a structured color distribution.<\/p>\n\n\n\n
Step 5 \u2014 One-to-One Tonal Mapping<\/h2>\n\n\n\n Instead of averaging everything into a single number, Shade Master:<\/p>\n\n\n\n
\nAligns sorted tooth palette with sorted shade palette<\/li>\n\n\n\n Performs one-to-one mapping<\/li>\n\n\n\n Calculates \u0394E for each paired color<\/li>\n\n\n\n Computes the average \u0394<\/li>\n<\/ul>\n\n\n\nThe shade with the lowest average Delta E<\/strong> wins.<\/p>\n\n\n\nNot the brightest.
Why This Matters Clinically<\/h1>\n\n\n\n For dentists:<\/p>\n\n\n\n
\nAccounts for cervical-to-incisal variation<\/li>\n\n\n\n Reduces operator bias<\/li>\n\n\n\n Uses perceptual color space (CIELAB)<\/li>\n\n\n\n Produces repeatable, objective comparisons<\/li>\n\n\n\n Improves communication with labs<\/li>\n<\/ul>\n\n\n\nIt transforms shade matching from a subjective judgment into a measurable comparison.<\/p>\n\n\n\n
Why This Matters Technically<\/h1>\n\n\n\n For developers:<\/p>\n\n\n\n
Shade Master demonstrates:<\/p>\n\n\n\n
![\"\"\/](\"https:\/\/raw.githubusercontent.com\/elselawi\/shade-master\/refs\/heads\/main\/alg.png\")
<\/figure>\n\n\n\n