detect.py

# import the necessary packages
from motiondetector import MotionDetector
import numpy as np
import argparse
import imutils
import cv2

# construct the argument parser and parse the arguments
ap = argparse.ArgumentParser()
ap.add_argument("-b", "--bounding-box", required=True, help="comma separated list of top, right, bottom, left coordinates of hand ROI")
ap.add_argument("-v", "--video", required=False, help="path to the (optional) video file")
args = vars(ap.parse_args())

# if a video path was not supplied, grab the reference to the webcam
if not args.get("video", False):
    camera = cv2.VideoCapture(0)

# otherwise, grab a reference to the video file
else:
    camera = cv2.VideoCapture(args["video"])

# unpack the hand ROI, then initialize the motion detector and the total number of
# frames read thus far
(top, right, bot, left) = np.int32(args["bounding_box"].split(","))
md = MotionDetector()
numFrames = 0

# keep looping
while True:
    # grab the current frame
    (grabbed, frame) = camera.read()

    # if we are viewing a video and we did not grab a frame, then we have reached the
    # end of the video
    if args.get("video") and not grabbed:
        break

    # resize the frame and flip it so the frame is no longer a mirror view
    frame = imutils.resize(frame, width=600)
    frame = cv2.flip(frame, 1)
    clone = frame.copy()
    (frameH, frameW) = frame.shape[:2]

    # extract the ROI, passing in right:left since the image is mirrored, then
    # blur it slightly
    roi = frame[top:bot, right:left]
    gray = cv2.cvtColor(roi, cv2.COLOR_BGR2GRAY)
    gray = cv2.GaussianBlur(gray, (9, 9), 0)


    # if we have not reached 32 initial frames, then calibrate the skin detector
    if numFrames < 32:
        md.update(gray)

    # otherwise, detect skin in the ROI
    else:
        # detect motion (i.e., skin) in the image
        skin = md.detect(gray)

        # check to see if skin has been detected
        if skin is not None:
            # unpack the tuple and draw the contours on the image
            (thresh, c) = skin
            cv2.drawContours(clone, [c + (right, top)], -1, (0, 255, 0), 2)
            cv2.imshow("Thresh", thresh)

    # draw the hand ROI and increment the number of processed frames
    cv2.rectangle(clone, (left, top), (right, bot), (0, 0, 255), 2)
    numFrames += 1


    # show the frame to our screen
    cv2.imshow("Frame", clone)
    key = cv2.waitKey(1) & 0xFF

    # if the 'q' key is pressed, stop the loop
    if key == ord("q"):
        break

# clean up the camera and close any open windows
camera.release()
cv2.destroyAllWindows()