Image editing with instance segmentation and image inpainting
The objective of this project is to create a robust Image Editing application that utilizes two powerful image processes: Instance Segmentation and Image Inpainting. By introducing these processes into the application, the application can decompose a given image into individual object and background...
Saved in:
Main Author: | |
---|---|
Other Authors: | |
Format: | Final Year Project |
Language: | English |
Published: |
Nanyang Technological University
2023
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/166892 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
Summary: | The objective of this project is to create a robust Image Editing application that utilizes two powerful image processes: Instance Segmentation and Image Inpainting. By introducing these processes into the application, the application can decompose a given image into individual object and background layers, allowing for a wide range of possibilities in terms of image manipulation.
Named ImageLab, the application has undergone multiple ideation and prototyping processes, and design considerations were also made to enhance the user experience. Product requirements were carefully crafted and referred to during the development of ImageLab to ensure traceability and consistency throughout the project. Two pre-trained models were used in this project: Mask R-CNN and PICNet for instance segmentation and image inpainting tasks, respectively. The performance of the different PICNet models were compared through experiments to identify the optimal one. To also ensure that ImageLab meets the requirements stated in this document, different test cases have been crafted and evaluated for correctness.
This report discusses the different design considerations made during the development of ImageLab, as well as the implementation and approach for each feature. The report also documents down the results of the various test cases as mentioned above, in order to test the functionality of the application. Through reading the report, one will also be able to gain a clearer understanding of how to use ImageLab and what they can expect from the application. |
---|