A Kinect recorder that captures and saves RGB image, depth(point cloud), depth frame mapped into color frame and skeleton position at 30FPS. Color images are stored in .jpg while point cloud is stored in .pcd and skeleton poistions in .txt. For each frame, you will need about 435KB for color image, point cloud about 1.9MB, and depth in color image about 200KB. So please have enough disk space left. If you don't need any recording source, simply comment out the corresponding part and recompile. Mutiple threads are used and depending on your CPU power, you may need to increase or decrease the number of threads in line 413-427.
Resolution: RGB: 1920 x 1080 Point Cloud: N x 3 (depending on how many invalid values) Depth Frame Mapped into Color: 1920 x 1080
Also, please set your own paths for all the depended libraries in the project setting since your installation path may differ from mine.
Dependencies: OpenCV Kinect for Windows SDK 2.0 Point Cloud Library (for point cloud only)
Note: To save point cloud, you can also use .txt file but it will be around 4MB for each frame. You may also replace the part where I save the image with OpenCV to make this software less dependent on other libraries.
Additional Reading: