Applying Stereo Vision to 3D Depth Perception in Embedded Systems
Depth information can help overcome a variety of Computer Vision challenges, such as detecting obstacles in automotive applications or interpreting gestures for smart user interfaces. We explored the advantages and limitations of a stereo vision depth camera, and demonstrated its feasibility on a range of embedded processors.
Depth information can help overcome a variety of Computer Vision challenges, such as detecting obstacles in automotive applications or interpreting gestures for smart user interfaces. We explored the advantages and limitations of a stereo vision depth camera, and demonstrated its feasibility on a range of embedded processors.
EVA video interview with hands-on demonstration of TI Stereo Module (TISMO) prototype on the C6678 Keystone DSP, wherein we highlight 1080p HD resolution disparity computation in real-time at 10W power consumption. We show-case how depth information can help detect obstacles for automotive safety, and share our experiences on TI System-On-Chips ranging from OMAP and DaVinci application processors all the way to Keystone multicore DSP devices. Full video available, © 2013 EVA.
Embedded 3D Stereo Vision: How It Works, How to Implement It, and How to Use It , a tutorial presented at the April 2013 Embedded Vision Summit. We demonstrate TI Stereo Module (TISMO), a DSP-optimized SW solution. We show how stereo depth information can help in various computer vision problems, including motion detection for video security, and obstacle detection for automotive and industrial safety. Full video and slides available, © 2013 EVA.
Demonstrating the TI Stereo Module (TISMO) on the DM8127 Stereo IP Network Camera by Leopard Imaging at the April 2013 Embedded Vision Summit. Contributors include Darnell Moore.
Demonstrating the TI Stereo Module (TISMO) on the OMAP L-138 Stereo Vision Development Kit with Tom Catalino from Critical Link at Photonics West (February 2013). Contributors include Peter Barnum.