Super deep 3D images from a 3D omnifocus video camera

Area of Science:

• Optics and Photonics
• Computer Vision
• Human Vision

Background:

• Stereographic image pairs are used for 3D image generation.
• Deep depth of field in original scenes improves 3D depth perception.
• Omnifocus video cameras offer unlimited depth of field, ensuring consistent focus.

Purpose of the Study:

• To investigate the use of omnifocus video cameras for real-time super deep stereoscopic image pair acquisition.
• To address the challenges posed by increased perspective shift in deep depth of field stereoscopic imaging.
• To develop a method for reducing perspective shift without compromising depth of field.

Main Methods:

• Utilizing an omnifocus video camera with an attachment to capture real-time stereoscopic video.
• Analyzing the relationship between depth of field and perspective image shift in stereoscopic pairs.
• Developing and implementing a technique to mitigate excessive perspective shift.

Main Results:

• Real-time super deep stereoscopic video image pairs were successfully obtained.
• A significant increase in perspective image shift was observed due to the deeper depth of field.
• A method was identified to reduce the perspective shift effectively.
• The developed method did not negatively impact the depth of field.

Conclusions:

• Omnifocus cameras enable enhanced stereoscopic 3D imaging with superior depth perception.
• Managing perspective shift is crucial for comfortable binocular fusion with deep depth of field stereoscopic video.
• The developed technique offers a practical solution for improving the quality of deep depth of field stereoscopic content.