Acquisition of 3-D data by focus sensing

Area of Science:

• Computer Vision
• Robotics
• Optical Sensing

Background:

• 3D shape reconstruction is crucial for robot vision and object recognition.
• Existing methods may lack speed or simplicity for real-time applications.

Purpose of the Study:

• To demonstrate a novel range sensing technique for acquiring the 3D shape of diffusely reflecting objects.
• To evaluate the technique's suitability for robot vision tasks requiring scene overviews.

Main Methods:

• Sequential depth sensing based on structured illumination.
• Focus sensing using a TV camera and analog electronics.
• Real-time processing of focus series for depth information.

Main Results:

• Successfully determined the 3D shape of objects.
• Achieved real-time performance for scene overview.
• Depth resolution was found to be limited.

Conclusions:

• The demonstrated technique is simple, rapid, and effective for obtaining 3D object shape overviews.
• It is well-suited for robot vision applications where speed is prioritized over high depth resolution.