Computer Vision: Extracting Data from the Visual World
1 like396 views
The document discusses computer vision and related technologies presented by Componica, LLC. It covers topics like facial detection, facial recognition, image registration, augmented reality, QR codes, and optical character recognition. Examples of applications are identifying diseases with smartphones, emotion detection, and automated manufacturing inspection. The conclusion emphasizes opportunities for computer visioning technology in society.
Computer Vision: Extracting Data from the Visual World
- 1. Computer Vision: Extracting Data from the Visual World A Brief Example... ! Steven Mitchell, Ph.D. Componica, LLC
- 2. About us. Componica, LLC (http://www.componica.com/) Strong Background in Computer Vision Copyright 2011 - Componica, LLC (http://www.componica.com/)
- 3. About us. Componica seamlessly combines the worlds of machine learning, computer visioning & mobile development & applying the latest in visionary technology to the world of mobile media. Words for Spanish / Russian / French Copyright 2011 - Componica, LLC (http://www.componica.com/)
- 4. Copyright 2011 - Componica, LLC (http://www.componica.com/) Why is computer vision relevant? How do these things work? Should I be concernd?
- 5. Copyright 2011 - Componica, LLC (http://www.componica.com/) In this slideshow: Facial Detection - Find me a face. Facial Recognition - Who’s face is it? Image Registration - Aligning pictures together. ...which leads to augmented reality. QR Codes - They’re everywhere. Optical Character Recognition - Reading Stuff.
- 6. Copyright 2011 - Componica, LLC (http://www.componica.com/) Face Detection This is NOT facial recognition. Developed by Viola / Jones in 2000. Major break-thru in image recognition...this was not possible prior. How much does a cow weigh? An army of simple face detectors. "Robust Real-time Object Detection"! Paul Viola and Michael Jones
- 7. Copyright 2011 - Componica, LLC (http://www.componica.com/) BTW, It’s how the Kinect sees people.
- 8. Copyright 2011 - Componica, LLC (http://www.componica.com/) BTW, It’s how the Kinect sees people. "Real-Time Human Pose Recognition in Parts from Single Depth Images"! Shotton, Fitzgibbon, Cook, Sharp, Finocchio, Moore, Kipman, Blake! Microsoft Research Cambridge & Xbox Incubation
- 9. Copyright 2011 - Componica, LLC (http://www.componica.com/) Facial Recognition Remove effects caused by lighting and perspective. After you find a face, reduce it to numbers. "Statistical Models of Appearance for Computer Vision"! T.F. Cootes and C.J.Taylor
- 10. Copyright 2011 - Componica, LLC (http://www.componica.com/) Facial Recognition Let’s mix some paint... Comparing numbers in hyperspace k-Nearest Neighbor, Wikipedia
- 11. Copyright 2011 - Componica, LLC (http://www.componica.com/) The most common way to register images. Find the most interesting points on the two images. Compare all the interesting points from one image to the other forming matching pairs of points between images. Image Registration - Interesting Points
- 12. Copyright 2011 - Componica, LLC (http://www.componica.com/) Augmented Reality FAST interest point detection 0.55ms Building query bit masks 0.12ms Matching into database 0.35ms Robust pose estimation 0.1ms Total frame time 1.12ms Table 1. Timings for the stages of our approach on a dataset with images taken from within the range of trained viewpoints. Figure 5. Increasing the range of viewpoint bins in the training set allows more viewpoint invariance to be added in a straightforward manner. gests that the bit count dissimilarity score provides a reason- able way of scoring matches. To confirm this we computed the average number of inlier and outlier matches over all of the frames in the two sequences, and plotted these against the dissimilarity score obtained for the match in Figure 4. For the sequence on the left where the viewpoints are in- cluded in the training set many good matches are found in Once you have correspondence, you can compute 3D geometry. http://mi.eng.cam.ac.uk/~er258/work/fast.html http://nghiaho.com
- 13. Copyright 2011 - Componica, LLC (http://www.componica.com/) QR Codes http://en.wikipedia.org/wiki/QR_Code ! "Quick Response code" invented by Toyota subsidiary Denso Wave in 1994. Open License Up to 2.5K of data Error Correction Easy to read and generate: ZXing library
- 14. Copyright 2011 - Componica, LLC (http://www.componica.com/) Optical Character Recognition iPhone 4th Gen iPod Touch 4th Gen
- 15. Copyright 2011 - Componica, LLC (http://www.componica.com/) Optical Character Recognition
- 16. Copyright 2011 - Componica, LLC (http://www.componica.com/) Commentary Ubiquitous Surveillance...extreme dislike. Birthday Paradox...The probability that, in a set of n randomly chosen people, some pair of them will have the same birthday.
- 17. Copyright 2011 - Componica, LLC (http://www.componica.com/) Commentary Video Cameras may fit the criteria of legally blind.
- 18. Copyright 2011 - Componica, LLC (http://www.componica.com/) Computer visioning technology and society: opportunities, possibilities: Smartphones that ID diseases, plants, insects. Robotic lawnmowers that don’t run over the neighbor’s cat. Computers that judge emotions by reading your face. Keyless entry based on face, iris. Automated inspection of manufactured parts. Conclusion