obendidi / Tracking-with-darkflow

Real-time people Multitracker using YOLO v2 and deep_sort with tensorflow
GNU General Public License v3.0
524 stars 176 forks source link

Using tracking with darkflow from another python application #56

Open mandeepubc opened 6 years ago

mandeepubc commented 6 years ago

Hi, I am trying to use this application from another application in python.


import cv2 from darkflow.darkflow.defaults import argHandler #Import the default arguments from darkflow.darkflow.net.build import TFNet import numpy as np import time

option = { 'model': 'darkflow/cfg/yolo-voc.2.0.cfg', 'load': 'darkflow/bin/yolo-voc_26600.weights', 'threshold': 0.15, 'gpu': 0.5, 'track': True, 'trackObj': ["person"], 'saveVideo': False, 'BK_MOG':True, 'tracker': "deep_sort", 'skip': 0, 'csv': True, 'display':True }

tfnet = TFNet(option)

capture = cv2.VideoCapture(0) # using webcamera colors = [tuple(255 * np.random.rand(3)) for i in range(5)]

while (capture.isOpened()): stime = time.time() ret, frame = capture.read() contours_now = [] if ret: results = tfnet.return_predict(frame)

    for color, result in zip(colors, results):
        tl = (result['topleft']['x'], result['topleft']['y'])
        br = (result['bottomright']['x'], result['bottomright']['y'])
        label = result['label']
        frame = cv2.rectangle(frame, tl, br, color, 7)
        frame = cv2.putText(frame, label, tl, cv2.FONT_HERSHEY_COMPLEX, 1, (0, 0, 0), 2)

    cv2.imshow('frame', frame)
    print('FPS {:.1f}'.format(1 / (time.time() - stime)))
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break
else:
    capture.release()
    cv2.destroyAllWindows()
    break

After results = tfnet.return_predict(frame), I want to feed this data to a tracking algorithm (deep_sort), so that I can get results with trackid. How can I do that? Thanks

monoloxo commented 4 years ago

可以看一下net/yolov2/predict.py中的相关函数