This document outlines a project to develop a resolution-independent 2D cartoon video conversion codec. The project aims to convert raster image frames in videos to vector image frames to overcome resolution dependency issues. Key aspects of the project include segmenting scenes, extracting and vectorizing backgrounds, extracting and tracking foreground objects, and reproducing the animation using the vectorized components. The project is implemented using OpenCV, C++, and vectorization tools like Potrace. Experiments demonstrate techniques for scene detection, background subtraction, object tracking, and animation reproduction. Further work is planned to automate color selection and optimize vectorization parameters.
5. Introduction
❏ Most of thevideos are in good quality in specificresolution and can be
clearlyviewedthrough thespecific resolution.
❏ When a lowquality video isplayed in high resolutionframes are blurred
and pixelated.
❏ Our aim is tocome up witha solution that would overcome resolution
dependency of videos
6. Background&Motivation
❏ High quality videois always in demand
❏ Bandwidth and storage problems
❏ The aim is toimprove thevisual appearance of the video
❏ Can use videoprocessing/Image processing technique
7. Problemin brief
❏ When resolution of a videois increased, most of thedigitalvideos get
blurred and pixelated
9. Aim
❏ Provide an effectiveapplication toconvert 2Dcartoon videos
independent on resolution
10. Objective
❏ Learn about video processing and imageprocessing
❏ Study about theobjectdetection,identificationand itsproblems
❏ Studyabout motion tracking algorithms
❏ Studytheproblems in converting rasterimage tovectorimage
❏ Studythealgorithms in scenesegmentation
❏ Studyabout animation and transformation of images
❏ Designand develop thesystemtoconvert 2D cartoon videos toany
resolution withoutaffecting thequality of a video.
11. Scope
❏ Wehave limitedour scope to2D cartoons.
❏ support for abrupt scenechanges
❏ simple user assistance needed for track themotionof foreground objects
15. List
❏ The modules above willmaintain and refer thefollowing liststhroughout
theprogram, which shall be common and available toallmodules. The
listswillmaintain referencesand cross referencestodata itemsthatare
processed by thesemodules.
● Scenelist
● Background list
● Character list
● Vectorbackground list
● Vectorcharacterlist
● Animation List
26. ObjectMotionPathDetection
Main task of themodule is monitoring themovementof thecharacters and
track thepath of each moving character.As a resultof this animation list
created.
❏ color of object istaken as a user input
❏ convert BGRtoHSV color space values and assign minimum and
maximum HSV values
❏ Threshold alltheframes totheHSV color
❏ filterobject and track x,y coordinates
29. Reproducing
The main task of this module is toreproduce theoutput video.
❏ Todo that thismodules takes allthelist createdby other modules and
motion path of each object.
❏ Place vectorbackground on theoutput image
❏ Place objectson thebackground on therelevant points.
❏ The final output can be given as vectorframes orreproduces as video
withraster frames scales tothe neededresolution