It is the repository of the final challenge project for Computer Vision (SS23) at Technical University Munich.
Lecturer: Prof. Dr.-Ing. Klaus Diepold
Assistant: Luca Sacchetto, Sven Gronauer
Team Member:
- Feng Xu
- Haoqing Huang
- Lei Zhang
- Long Zhao
- Yuyang Zhang
This project develops an application that allows customers to check the interior of a 3D scene of their own apartment. The customer is able to easily upload their own captured images of the room into the application and receive a 3D model of the interior space.
💻 : work for Windows
🚀 : MATLAB Version: at least MATLAB R2023a and Toolbox Computer Vision Toolbox must be installed
⌨️ : GUI develop kit: APP designer
First, open the CV_challenge.mlapp in the current folder.
Next, click the Run button. Alternatively, you can click Code View to see the code of GUI.
The GUI looks like this.
Select the folder of the images you want to upload. Also select the folder containing the camera paramters. Camera parameters will be displayed automatically. Then click Confirm. The first two lights should be green and then you can click Run to start the computation.
After running, the point cloud and 3D model will be displayed. Figure 1 is the point cloud without clustering and figure 2 with clustering.
3D model is shown in the center of the interface. You can drag to view it from different positions. On the right side of the 3D model, volume of each object and the distance between their center and original point are shown. If you want to try a new image, please use the Reset button.
