lzhbrian / Palm-Live-Detection

Palm Live Detection (Project of Digital Image Processing Course)

Palm Live Detection

Brief

To detect if a palm is alive or is just a picture, implementing using OpenCV.

This system is a course project of Digital Image Processing.

brief steps:

1. Find palm image pixels using OTSU algorithm
2. Find Contour of the pixels
3. Get finger end vertices
4. Monitoring the relative distance between finger end vertices.
   1. If changed to a threshold -> ALIVE

Usage

g++ source/hand_reco_lzhbrian.cpp \
-I/path/to/opencv3/include \
-L/path/to/opencv3/lib \
-lopencv_core -lopencv_highgui \
-lopencv_videoio -lopencv_objdetect \
-lopencv_imgproc -lopencv_imgcodecs \
-o hand_reco # To compile

./hand_reco # To run

Demo

• This system detects the relative distance of finger end vertices, so moving forward or backward won't trigger anything. (as shown below)

• If you stretch your fingers, which results in change of relative distance between finger end vertices, the system will consider it as a LIVE PALM. (as shown below)