README.md

AirInterview

This document contains the assignment for role of computer vision position.

Assuming you know the basic knowledge about C++, CMake and Docker, this will only take you at most 2 hours to finish the basic requirement.

Requirements

Fork this project and create a pull request once finished the assignment.

Goal

1. Verify candidate's ability to work with point cloud data
2. Verify candidate's ability to read documentation and implementation skill
3. Verify candidate's ability to write clean and testable code

Assignment

Cylinder model segmentation

By using table_scene_mug_stereo_textured.pcd, find the cylinder object inside the scene.

Input

• table_scene_mug_stereo_textured.pcd

Output

• Cylinder inlier point cloud file

Bonus(Optional)

1. Visualization screenshot or video

2. Render the final result in web

3. TDD(test driven development) approach

4. Good documentation habit including code comment, project readme.

Recommendation

The following is the recomendded way to solve this assignment but it is not limited within this scope. Feel free to use your own way to create a solution. That will be a bonus

1. Cylinder model segmentation tutorial in PCL

2. Dev environment setup ---

   • Using PCL docker image provided by Airsquire
   • Or following the PCL official tutorial

3. Testing framework -- Google Test. You can use my blog post as a reference.

4. Web rendering -- Three.js