def run_darkflow():
options = {
"model": "./cfg/v1.1/tiny-yolov1.cfg",
"load": "./bin/tiny-yolo-v1.1.weights",
"threshold": 0.1,
"demo": "camera",
}
tfnet = TFNet(options)
return tfnet.camera(pl1, pl2, t1, t2)
class YOLOv2Subject(Subject):
def __init__(self, FLAGS):
self.observers = []
self.yoloNet = TFNet(FLAGS)
def attach(self, observer):
self.observers.append(observer)
def detach(self, observer):
self.observers.remove(observer)
def notify(self, result):
print("YOLOv2Subject: update people detection")
for observer in self.observers:
observer.update(result)
def detectImage(self, image):
self.notify(self.yoloNet.return_predict(image))
def detectVideo(self):
self.yoloNet.camera()
def cliHandler(args):
FLAGS = argHandler()
FLAGS.setDefaults()
FLAGS.parseArgs(args)
# make sure all necessary dirs exist
def _get_dir(dirs):
for d in dirs:
this = os.path.abspath(os.path.join(os.path.curdir, d))
if not os.path.exists(this): os.makedirs(this)
_get_dir([
FLAGS.imgdir, FLAGS.binary, FLAGS.backup,
os.path.join(FLAGS.imgdir, 'out'), FLAGS.summary
])
# fix FLAGS.load to appropriate type
try:
FLAGS.load = int(FLAGS.load)
except:
pass
tfnet = TFNet(FLAGS)
if FLAGS.demo:
tfnet.camera()
exit('Demo stopped, exit.')
if FLAGS.train:
print('Enter training ...')
tfnet.train()
if not FLAGS.savepb:
exit('Training finished, exit.')
if FLAGS.savepb:
print('Rebuild a constant version ...')
tfnet.savepb()
exit('Done')
tfnet.predict()
from darkflow.net.build import TFNet
options = {
"model": "cfg/yolo-voc-hyperion.cfg",
"load": "weights/yolo-voc_50000.weights",
"threshold": 0.1,
"labels": "image.names",
"demo": "camera",
"gpu": 0.5,
"address": 'localhost',
"port": 48051,
"UDP": True
}
tfnet = TFNet(options)
#imgcv = cv2.imread("./darkflow/sample_img/sample_dog.jpg")
#result = tfnet.return_predict(imgcv)
tfnet.camera()
exit("Demo Stopped, exiting.")
class yolo(object):
modeldir = yolo_dir
parentdir = cwd
def __init__(self,):
print("yolo model created")
def load(self,model='tiny-yolo',threshold=0.25,gpu=0):
os.chdir(yolo_dir)
model_list =['tiny-yolo','tiny-yolo-voc','yolo']
model_custom_list=['yolo-hand','tiny-yolo-hand']
if model in model_list:
self.options = {"model": "cfg/"+model+".cfg", "load": "bin/"+model+".weights", "threshold": threshold,'gpu':gpu}
elif model in model_custom_list:
self.options = {"model": "cfg/"+model+".cfg", "load": 100,"threshold": threshold,'gpu':gpu}
else:
raise ValueError('model passed in is not in supported model list. Please pass in correct model.(tiny-yolo;tiny-yolo-voc,yolo)')
self.tfnet = TFNet(self.options)
os.chdir(self.parentdir)
#print(self.parentdir)
return None
def predict(self,img_path):
imgcv = cv2.imread(img_path,1)
result = self.tfnet.return_predict(imgcv)
return result
def predict_imgcv(self,imgcv):
result = self.tfnet.return_predict(imgcv)
return result
def predict_imgcv_list(self,imgcvs,threshold=0.3):
buffer_inp = list()
buffer_pre = list()
boxex_list = list()
for fr in imgcvs:
preprocessed = self.tfnet.framework.preprocess(fr)
buffer_pre.append(preprocessed)
buffer_inp.append(fr)
feed_dict = {self.tfnet.inp: buffer_pre}
net_out = self.tfnet.sess.run(self.tfnet.out, feed_dict)
for im,single_out in zip(buffer_inp, net_out):
h, w, _ = im.shape
boxes = self.tfnet.framework.findboxes(single_out)
boxesInfo = list()
for box in boxes:
tmpBox = self.tfnet.framework.process_box(box, h, w, threshold)
if tmpBox is None:
continue
boxesInfo.append({
"label": tmpBox[4],
"confidence": tmpBox[6],
"topleft": {
"x": tmpBox[0],
"y": tmpBox[2]},
"bottomright": {
"x": tmpBox[1],
"y": tmpBox[3]}
})
## add processed boxex into a list
boxex_list.append(boxesInfo)
return boxex_list
def draw_save_pic(self,img_path,out_path):
imgcv = cv2.imread(img_path,1)
result = self.tfnet.return_predict(imgcv)
draw_box(result,imgcv)
cv2.imwrite( os.path.join(out_path), imgcv);
return imgcv
def draw_return_pic(self,frame):
buffer_inp = list()
buffer_pre = list()
results = list()
for fr in frame:
preprocessed = self.tfnet.framework.preprocess(fr)
buffer_pre.append(preprocessed)
buffer_inp.append(fr)
feed_dict = {self.tfnet.inp: buffer_pre}
net_out = self.tfnet.sess.run(self.tfnet.out, feed_dict)
for img, single_out in zip(buffer_inp, net_out):
postprocessed = self.tfnet.framework.postprocess(single_out, img, False)
results.append(postprocessed)
return results
def demo(self,option='camera'):
self.tfnet.FLAGS.demo = option
self.tfnet.camera()
return None
def video_return_json(self,option):
camera = cv2.VideoCapture(0) ## 0 is the default camera
while camera.isOpened():
_,frame = camera.read()
if frame is None:
print('\n End of Video')
result = self.tfnet.return_predict(frame)
sys.stdout.write('%s\r' % result[0])
sys.stdout.flush()
#cv2.imshow('',frame)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
camera.release()
cv2.destroyAllWindows()
def start_camera():
from darkflow.net.build import TFNet
tfnet = TFNet(FLAGS)
tfnet.camera()
exit('Demo stopped, exit.')