Driver Assistant Solutions with
summary
Open-source Android prototype called pony-drive repo bit.ly/pony-drive shows computer vision driver assistance built with OpenCV on Android, using ROI + resize (to ~25%), grayscale, Gaussian blur, thresholding, Canny edges and Hough-style lane grouping, ORB/BRISK/BRIEF/FREAK feature descriptors, and simple traffic-light detection via minMaxLoc + circle drawing. Demo runs on-device but feels choppy; moving performance into C++ native -> 90% fewer JNI calls yields much faster processing. Hard parts: camera orientation, device variation, and plate recognition failing where lane detection succeeds.
AI-written summary · about
chapters (10)
- 0:00Introduction and background
- 1:57What is computer vision
- 6:58Image processing basics
- 13:43Feature detection demo
- 17:26OpenCV library overview
- 21:38Integrating OpenCV in Android
- 27:47Native C++ development in Android
- 33:24Traffic light detection
- 35:59Driving lane detection
- 39:53Q&A session
related talks
26mSelf Driving cars with Opencv
30mMobile applications that can see
41mAI applications in autonomous driving
06mInSight: Monitoring the state of the driver in low-light using smartphones
47mAn unexpected journey into computer vision, or There and Back Again
49mBuilding a self driving RC car
54mSurveillance Detection Scout – Your Lookout on Autopilot
27mAutomated Python Powered Turret