1 / 11

Multi-Cam Video Stitching

Multi-Cam Video Stitching. Panoramic Video Option During a Skype Call. Goal. Integrate a panoramic video feature into Multicam Feature takes the inputs of multiple webcams and stitches them together Output is produced in real time Works using hardware of the average user. Applications.

faolan
Télécharger la présentation

Multi-Cam Video Stitching

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Multi-Cam Video Stitching Panoramic Video Option During a Skype Call

  2. Goal • Integrate a panoramic video feature into Multicam • Feature takes the inputs of multiple webcams and stitches them together • Output is produced in real time • Works using hardware of the average user

  3. Applications • Conference calls • Absentee lecture

  4. Background: Image Stitching • Homography • Computer Vision by R. Szeliski • Types of algorithms • Improving the image

  5. Background: Video Stitching • Efficient stitching made use of GPU • Good quality videos not real time • Good quality but required user input

  6. Our Approach and Research Question • OpenCV • Basic hardware components • Can a good panoramic video be stitched in real time during a Skype call with the equipment of the average person? • “good” • “real time” • “equipment of the average user”

  7. Progress: Preliminary work with OpenCV • Install OpenCV on Mac • Simple image stitch

  8. Progress: Simple Videos • Still camera videos • 640 x 480 resolution • Compute homography for each frame • Compute homography once • Results • Typical video has 24 FPS • Approach #1: .89 FPS • Approach #2: 2.12 FPS

  9. Progress: Change the Resolution

  10. Future Work: Speed • Create a mask of videos, only quarter-sized, calculate the homographies and then scale the homographies up to match real size. • Modify variables included within the Stitcher class. • Allow user to select position on slidebar to adjust precision and performance.

  11. Future Work: Quality • Exposure, Gain • Use built in methods in the stitcher class • Observe effects on time • Try videos in different light, observe effects of built in methods

More Related