Carleton University - School of Computer Science Honours Project
Summer 2021
Cel-Shaded Image Stylization
Micheal Hamon
SCS Honours Project Image
ABSTRACT
This project seeks to build a program which can use Image Stylization techniques to take an unedited digital photograph and automatically convert it to a cel-shaded style. The intended primary use case for this program are small indie developers who currently often use digital photographs directly as backgrounds even when the rest of their game assets are drawn in the cel-shaded style as producing a large number of drawn backgrounds is not within their budget. The process to achieve this will primarily consist of two portions, Colour Quantization to process the image into a stylized representation of itself, and Edge Detection which will be used to give the image outlines for the purpose of giving definition to the individual elements existing within the image. This project hopes to demonstrate that using only this program and the digital image a developer would have already been planning to use with their work a cel-shaded stylized image can be created which will better mesh with their work’s artistic style.