top of page

FUNG JUN YIN

Robots-Square.jpg

My passion for techs can be traced back to my SPM days. As a student with additional ICT subject, I have been exposed to techs unquestionably. Few years have passed and I have successfully enrolled into USM as a Computer Science Student. Throughout this whole undergraduate journey, I have learnt plenty of skills and may they be technological skills, communication skills or problem-solving skills, I believe they will help me to grow into a better person. For further information, I am experienced in Docker, Azure Devops, AWS EC2 and et cetera. Feel free to contact me through fungjunyin123@gmail.com or 010-3722933

Matrix No:

137070

Student Email: 

Supervisor:

Dr. Yap Fa Toh

Supervisor Email: 

linkedin_edited.png
—Pngtree—instagram icon instagram logo_3
Project

SC019

photo manipulation mobile app

Nowadays, there is indeed a lot of photo editing app which can be found in the Android market but a lot of them have the tendency to rely on another app to do certain task. For instance, most of photo editing apps are not capable to allow user to change the photo background to another background unless switching to use background changing app. Moreover, most of the background changing app is not able to run without internet connection. Thus, this project is proposed to address these issues. This proposed project is a photo editing app developed for Android using Java and is set to target user from all age group. This app is equipped with standard photo editing features and photo background changing feature. The standard photo editing will be composed of text insertion, filter insertion and photo enhancement tool for brightness, contrast and saturation. Whereas for photo background changing feature, this app will provide an automatic background remover feature for any human photo. The purpose of removing background is to extract the foreground so user can input a new photo as background. The process of removing background is done automatically by using image segmentation. The image segmentation is powered by machine learning for accurate human detection. Furthermore, this app also provides a new feature to merge multiple foregrounds as one foreground. This feature is designed with the intention to allow users to create a foreground that consists of multiple persons instead of one person only. There are also features for restoring and removing foreground so that the users can manually fine tune the foreground. Lastly, this whole app is not internet reliant and capable to run any process on-device.


Gallery

Gallery

Demo

Demo

Other Project

bottom of page