This project converts real photos into cartoon images using OpenCV python. This is mainly done through applying edge detection mechanism to produce bold silhouettes and some filters such as the bilateral filter to reduce the color palette and median blur filter to remove the noise. The dataset used for testing images is a subset from “1 Million Fake Faces” dataset on Kaggle and contains 1029 images. Since the cartoon effect is somehow subjective, the parameters controlling the resulting images were tuned to produce the maximum possible cartoon effect.
• Link for dataset: shorturl.at/clTX7
• Link for sample results: shorturl.at/hqCWY