smarthome-presence-detect/camera/person_detection.py

#!/usr/bin/env python

from imutils.video import VideoStream
from imutils.video import FPS
import argparse
import imutils
import cv2
from datetime import datetime, time
import numpy as np
import time as time2

VISUAL_DEBUG=True

def getArgs():
    """ Arguments """
    ap = argparse.ArgumentParser()
    ap.add_argument("-v", "--video", help="path to the video file")
    ap.add_argument("-a", "--min-area", type=int, default=500, help="minimum area size")
    ap.add_argument("-t", "--tracker", type=str, default="csrt", help="OpenCV object tracker type")
    return vars(ap.parse_args())


def main():
    args = getArgs()

    # if the video argument is None, then the code will read from webcam (work in progress)
    if args.get("video", None) is None:
        vs = VideoStream(src=0).start()
        time2.sleep(2.0)
    # otherwise, we are reading from a video file
    else:
        vs = cv2.VideoCapture(args["video"])

    cv2.namedWindow('Video stream', cv2.WINDOW_NORMAL)
    detector = DetectionFromFrame(args["min_area"], 0.8)
    while True:
        detector.currentFrame = vs.read()
        detector.currentFrame = detector.currentFrame if args.get("video", None) is None else detector.currentFrame[1]
        # if the frame can not be grabbed, then we have reached the end of the video
        if detector.currentFrame is None:
            break

        # resize the frame to 500
        detector.currentFrame = imutils.resize(detector.currentFrame, width=500)
        detector.framecounter+=1
        if detector.framecounter > 1:
            cnts = detector.prepareFrame()
            
            for c in cnts:
                boundRect = cv2.boundingRect(c)
                #(x, y, w, h) = cv2.boundingRect(c)
                #initBB2 =(x,y,w,h)

                prott1 = r'ML-Models/MobileNetSSD_deploy.prototxt'
                prott2 = r'ML-Models/MobileNetSSD_deploy.caffemodel'
                net = cv2.dnn.readNetFromCaffe(prott1, prott2)

                #trackbox = detector.currentFrame[y:y+h, x:x+w]boundRect[1]
                trackbox = detector.currentFrame[boundRect[1]:boundRect[1]+boundRect[3], 
                                                boundRect[0]:boundRect[0]+boundRect[2]]
                trackbox = cv2.resize(trackbox, (224, 224))
                #cv2.imshow('image',trackbox)
                
                blob = cv2.dnn.blobFromImage(cv2.resize(trackbox, (300, 300)),0.007843, (300, 300), 127.5)
                net.setInput(blob)
                detections = net.forward()

                for i in np.arange(0, detections.shape[2]):
                    detector.detectConfidentiallyPeople(i, detections, boundRect)
                cv2.rectangle(detector.currentFrame, (boundRect[0], boundRect[1]), 
                                (boundRect[0] + boundRect[2], boundRect[1] + boundRect[3]), (255, 255, 0), 1)

            
        # show the frame and record if the user presses a key
        cv2.imshow("Video stream", detector.currentFrame)
        key = cv2.waitKey(1) & 0xFF

        # if the `q` key is pressed, break from the lop
        if key == ord("q"):
            break
        if key == ord("d"):
            detector.firstFrame = None
        #detector.lastFrame = detector.currentFrame

    # finally, stop the camera/stream and close any open windows
    vs.stop() if args.get("video", None) is None else vs.release()
    cv2.destroyAllWindows()


class DetectionFromFrame:
    def __init__(self, min_size, confidence):
        self.min_size = min_size
        self.confidence_level = confidence

        self.firstFrame = None
        self.currentFrame = None
        
        self.initBB2 = None
        self.fps = None
        self.differ = None
        self.now = ''
        self.framecounter = 0
        self.people_count_total = 0


    def prepareFrame(self):
        gray = cv2.cvtColor(self.currentFrame, cv2.COLOR_BGR2GRAY)
        gray = cv2.GaussianBlur(gray, (21, 21), 0)

        # if the first frame is None, initialize it
        if self.firstFrame is None:
            self.firstFrame = gray
            return []

        # compute the absolute difference between the current frame and first frame
        frameDelta = cv2.absdiff(self.firstFrame, gray)
        thresh = cv2.threshold(frameDelta, 25, 255, cv2.THRESH_BINARY)[1]
        
        #debug
        """if VISUAL_DEBUG: 
            cv2.imshow("debug image", thresh)
            cv2.waitKey(0)
            cv2.destroyWindow("debug image")
            #cv2.destroyWindow("threshhold image")"""
        
        # dilate the thresholded image to fill in holes
        thresh = cv2.dilate(thresh, None, iterations=2)

        # find contours on thresholded image
        thresh = np.uint8(thresh)
        cnts, _  = cv2.findContours(thresh.copy(), cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)

        return cnts

    def detectConfidentiallyPeople(self, i, detections, boundRect):
        CLASSES = ["person"]

        COLORS = [0,255,0]
        #COLORS = np.random.uniform(0, 255, size=(len(CLASSES), 3))

        confidence = detections[0, 0, i, 2]

        if confidence > self.confidence_level:
            # extract the index of the class label from the `detections`, then compute the (x, y)-coordinates of
            # the bounding box for the object
            #idx = int(detections[0, 0, i, 1])
            box = detections[0, 0, i, 3:7] * np.array([boundRect[2], boundRect[3], boundRect[2], boundRect[3]])
            (startX, startY, endX, endY) = box.astype("int")
            # draw the prediction on the frame
            
            #label = "{}: {:.2f}%".format(CLASSES[idx], confidence * 100)
            label = "{}: {:.2f}%".format(CLASSES[0], confidence * 100)
            
            #cv2.rectangle(frame, (startX, startY), (endX, endY), COLORS[idx], 2)
            cv2.rectangle(self.currentFrame, (boundRect[0], boundRect[1]), 
                                (boundRect[0] + boundRect[2], boundRect[1] + boundRect[3]), (0,255,  0), 3)
            
            y = boundRect[1] - 15 if boundRect[1] - 15 > 15 else boundRect[1] + 15

            #cv2.putText(frame, label, (startX, y), cv2.FONT_HERSHEY_SIMPLEX, 0.5, COLORS[idx], 2)
            cv2.putText(self.currentFrame, label, (0, 15), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0,0,0), 1)
            #cv2.imshow("Video stream", self.currentFrame)
            #print("Person found")



if __name__ == "__main__":
    main()