Hi, I’m Anson. I am passionate about Computer Vision, 3D scanning, and reconstruction. With years of experience in this industry, I have honed my skills in transforming complex data into meaningful visual representations. My work reflects my commitment to innovation and excellence in these cutting-edge fields. I invite you to explore my portfolio and see how my expertise can bring ideas to life.
Technical Research and Project Management
Technical Development and Tool Building
Team Management and Collaboration
FINAL FANTASY XV POCKET EDITION - C#, Unity, Wwise
Unity Support Team (R&D Team) - C#, Unity
DTL(Web Game) - ActionScript, Adobe Flash
A simple app concept to log meals, portions, and notes daily, designed to make food tracking effortless and extendable to nutrition insights later.
“Record Your Love” is a dedicated tool for collectors to easily scan, organize, and manage their treasured items. Whether it’s books, limited edition figures, collectible wines, music albums, or even your favorite food list, you can quickly add them to your collection using barcode scanning or manual input. The app features built-in categorization and search functions, allowing you to find your items effortlessly, and also provides a QR code generation feature for easy sharing. With a clean and intuitive interface, you can conveniently record and manage your favorite things anytime, anywhere, ensuring that every cherished item is well-documented and never forgotten.
This flag quiz app blends education and entertainment, designed to help players learn and recognize flags from around the world. Built with Unity, the app supports both Chinese and English languages and offers five game modes, including Endless Challenge, Timed Mode, and Review Mode, catering to diverse preferences. On the technical side, it uses JSON to manage flag data, LeanTween for smooth UI animations and visual effects, and includes features like accuracy tracking and a personalized stats page to help players focus on improvement.The design emphasizes simplicity and intuitiveness, enhanced by dynamic button animations and sound feedback for an engaging experience. Future updates will introduce interactive map integration and additional game modes, further enriching the gameplay and providing a fun way to explore world flags.
Watch Video
I’m currently working on a project that allows users to simply input a stock ticker symbol and retrieve relevant stock information, including real-time prices, transaction volumes, price changes, opening prices, and KD values. The goal is to eventually integrate custom algorithms to filter out preferred stocks. The project is being developed in Python and runs on macOS.
View Project on GitHub
Because my child loves playing with light switches, I created a handheld toy that allows them to toggle different types of switches to individually control various light bulbs, satisfying their curiosity and desire to explore.
Watch Video
To help my pregnant wife quickly check which foods are not recommended during pregnancy, I developed a lightweight 2D app using Unity. The app allows for quick food safety checks, adding new food information to a personalized list, and indicating whether a food is safe to eat through an intuitive signal light.
View Project on GitHub
In my spare time, I designed and built an automated sensor trash can that opens when a user approaches. This project involved prototyping with Arduino, integrating sensors, and designing a compact mechanism. The trash can provides a touchless experience and is both practical and fun to use.
View Project on GitHub
This project was developed in collaboration with the water treatment plant to assess water quality by analyzing fish posture. Water from the source is introduced into two separate tanks, each containing a goldfish (Carassius auratus) that has adapted to the water conditions. Cameras are positioned at the front and side of each tank to monitor the fish. The system leverages background removal and dynamic object elimination to establish a clean background, combined with morphological processing to accurately capture the fish’s body region. An SVM model is then used to classify and recognize the fish’s orientation, enabling real-time statistical analysis and water quality assessment. If both tanks show signs of abnormal fish behavior, the system automatically sends a text message to notify maintenance personnel. The entire project was developed in C++, with multi-threading used to accelerate dual-channel video processing, and MFC employed for interface development.
Read Paper
This project is an autostereoscopic 3D display system that dynamically adjusts the displayed 3D content based on the user’s head position. The system first utilizes OpenCV for real-time face and head position detection, accurately capturing the user’s viewing angle. Bullet Physics SDK is then employed to simulate the physical behavior of objects in the scene, including realistic effects like gravity and collisions. Finally, OpenGL is used to render the corresponding 3D scene based on the detected viewing angle, ensuring a smooth and immersive 3D experience from different perspectives. The system delivers a glasses-free 3D display, making it ideal for interactive showcases and exhibitions.
Watch Video
This is an academic project focused on people counting using computer vision techniques. The system is equipped with a top-view camera that continuously monitors and tracks people flow in real time. The project uses OpenCV for background modeling and foreground detection, allowing the system to isolate moving targets. Morphological processing techniques are applied to accurately extract human contours for effective counting. To present the results intuitively, an interface was developed using MFC, which displays real-time counting results and processed images. This system is ideal for environments requiring crowd analytics, such as public transport stations, commercial spaces, or event venues, providing accurate and timely data support.
Watch Video
This project is a digit recognition system developed using MATLAB. It provides a handwriting interface where users can directly write the numbers they want to recognize. The system utilizes SVM (Support Vector Machine) for classification and integrates PCA for dimensionality reduction, ensuring high recognition accuracy while reducing data complexity. Through this process, the system effectively handles variations in handwritten inputs and quickly identifies the numbers, showcasing the application of machine learning in image recognition.
Watch Video
This project is a 3D capturing app that also includes 3D reconstruction features, initially developed to attract potential customers. I was fully responsible for the UI/UX design, as well as the development of the entire app. Users begin by registering or logging in and providing necessary personal information. After this, they proceed to the 3D capturing stage, where the app captures multiple angles of the user, which are then used to reconstruct a 3D object for analysis. The app also includes features for reviewing the generated 3D object and provides an onboarding page to guide users through the app. Additionally, the app offers a function that allows users to capture real-world objects by simply circling them, enabling 3D reconstruction directly on the mobile device.
Watch Video
This project was an app that I was fully responsible for developing at my previous company. Due to the need to display extensive 3D content, effects, and real-time computations, Unity was chosen as the development platform. The app features significant UI and UX design elements, along with various social media and social functionalities, such as sharing 3D content or screenshots. Additionally, the app includes the capability to download the latest 3D holograms from the server in real-time and display them directly within the app, offering a rich interactive experience.
Watch Video
These applications were developed with the aim of collaborating with potential clients, utilizing 3D visualization to replace traditional 2D display methods. The projects span industries such as sports, energy, and smart environments. Compared to 2D, 3D technology offers a more visual and comprehensive presentation of data and information. Additionally, these applications integrate 3D models with real-world terrain and statistical data, creating a more dynamic and interconnected visualization experience. I consolidated the results of these projects into a single showcase, utilizing various technologies such as UI/UX design, human-computer interaction, light and shadow switching, point cloud rendering, and a simulation system for crowd detection and statistics.
Watch Video
This project showcases my work in areas such as object detection, posture recognition, object classification, and the application of these technologies in services such as physical therapy and posture recognition. In these projects, I utilized a variety of tools and libraries, including Microsoft’s Azure Kinect Body Tracking SDK, OpenPose, MediaPipe, and extensive use of OpenCV libraries. For object detection, I employed YOLO (You Only Look Once) technology. These projects highlight the practical applications of these advanced technologies in real-world scenarios.
Watch Video
This project involved compiling and integrating videos of various AR demos I have worked on over the years into a comprehensive showcase, highlighting the evolution and application of AR technology. In these demos, object recognition technology progressed from traditional marker recognition, to image recognition, to 3D object recognition, and now the latest method using ARKit for spatial positioning. Most of the development was carried out on the Unity platform, utilizing AR Foundation, ARKit, and Vuforia. By consolidating these demo videos, I can systematically showcase how these AR technologies have been applied across different scenarios, emphasizing the diversity of the technology and its developmental journey.
Watch Video
This project involved developing a visualization pipeline and toolset in Unity, specifically designed to showcase motion capture and posture analysis techniques. The tool provides an intuitive way to display detailed information for each marked point within 3D holograms, including coordinates and velocity. These marked points represent key nodes in the motion, allowing for in-depth analysis and evaluation of the movements. The tool not only accurately presents the dynamic changes of these points but also helps users better understand complex motions and postures.
Watch Video
This project is an AR 3D object tracking app running on Android, developed using Unity. In the project, one mobile device serves as the server side to control the 3D content and synchronizes the control results to the client-side mobile device. The controls include displaying digital content and locking onto interactive objects. The virtual 3D content is positioned by tracking real-world 3D objects, enabling a seamless integration between the virtual and real environments.
Watch Video
This project focuses on real-time people tracking and crowd monitoring using LiDAR technology. LiDAR sensors are installed at key street entrances to detect and track pedestrian flow for crowd control. The system is integrated with a Unity-based interface, providing a user-friendly way to monitor crowd density and visualize 3D crowd movement. The data is continuously sent to AWS for statistical analysis to help manage crowd flow effectively. During the pandemic, this system proved particularly useful for controlling the number of people entering and exiting the area and issuing alerts when necessary. The project also received the Global ICT Excellence Award.
Watch Video Read NewsGPA: 4.33
In my free time, I love reading, watching movies, and catching up on TV series—they help me unwind and enjoy some quiet moments. Outdoors, I enjoy cycling and hiking, taking in the scenery, and snapping a few photos along the way. These activities keep me balanced and make life more enjoyable.
Besides reading and outdoor activities, I enjoy getting hands-on with fun DIY projects. I once made a simple automatic sensor trash can that opens when you approach it—really handy! I also created some interactive toys for my son, like a light that he can control with his hands. He loves playing with them. These little projects add some fun to my busy days and give my son some cool toys to enjoy.
I enjoy capturing life’s moments with my camera, whether it’s the warmth of everyday scenes or a beautiful view I stumble upon during a trip. Photography, for me, is a way to freeze those touching moments and appreciate the little things in life.
.