sanya-mahajan/hand_vid_effect

HAND CONTROLLED WEBCAM FILTERS

use your fingers to change the filter applied to live webcam. implemented using openCV (python) and python libraries - numPy and mediapipe for real time detection of hand.

the user can raise the number of fingers depending on the required effect.

mediapipe is used to determine a hand object. The model is already trained to determine location of the hand and different landmarks on the fingers. There are mainly 21 points which are indicated by filled dots and can be used to study the action performed.Then we draw lines to connect the landmarks.

a counter is maintained to track the number of fingers raised and accordingly the corresponding effect option is chosen.

the options are as follows:

the inbuilt functions of openCV are used to apply the filter on the captured frame and the result is seen in real time.