CARTONIZATION OF IMAGES AND VIDEOS USING GAN AND OPEN CV

Abstract

Cartoonization is a creative transformation of real-world images or videos into cartoon-style visuals. It has applications in social media filters, animated content creation, and digital art. This project proposes a system that uses Generative Adversarial Networks (GANs) in combination with OpenCV to cartoonize both static images and dynamic videos. While OpenCV provides image processing utilities such as edge detection and color smoothing, GANs—especially models like CartoonGAN and CycleGAN—are trained to learn the artistic style of cartoons and apply it to real images. The goal is to produce high-quality, real-time cartoonized outputs while maintaining the structure and integrity of the original input