def webreg(userexcel):
args = {
"east": "frozen_east_text_detection.pb",
"min_confidence": 0.5,
"width": 320,
"height": 320,
"padding": 0.0,
"webcam": "", # path of webcam
"excel": userexcel,
}
(W, H) = (None, None) # actual dimensions of image 720x460 etc
(newW, newH) = (args["width"], args["height"]) # required dimewnsion
(rW, rH) = (None, None) # ration of both
layerNames = [
"feature_fusion/Conv_7/Sigmoid", # Scores - probability
"feature_fusion/concat_3"
] # geometry - dimensions of the bounding box
print("loading EAST text detector...")
net = cv2.dnn.readNet(args["east"])
# if no webcam path, grabbing the reference to the web cam
print("[INFO] starting webcam stream...")
vs = WebcamVideoStream(src=0).start() # 0 for default webcam
# time.sleep(1.0)
fnumber = -10
vfname = []
T = []
predictedTexts = []
fps = FPS().start()
while True:
if args["webcam"]:
fnumber += 10
vs.set(cv2.CAP_PROP_POS_FRAMES, fnumber)
frame = vs.read()
frame = frame[1] if args.get("webcam", False) else frame
# check to see if we have reached the end of the stream
if frame is None:
break
# resize the frame maintain aspect ratio
frame = imutils.resize(frame, width=1000)
orig = frame.copy()
if W is None or H is None:
(H, W) = frame.shape[:2] # actual size
rW = W / float(newW)
rH = H / float(newH)
# resize the frame
frame = cv2.resize(frame, (newW, newH))
# construct a blob
blob = cv2.dnn.blobFromImage(frame,
1.0, (newW, newH),
(123.68, 116.78, 103.94),
swapRB=True,
crop=False)
net.setInput(blob)
(scores, geometry) = net.forward(layerNames)
# decode the predictions obtaining probabilites and position of box
(rects, confidences) = decode_predictions_video(scores, geometry, args)
boxes = non_max_suppression(np.array(rects), probs=confidences)
pytesseract.pytesseract.tesseract_cmd = r'C:\Program Files\Tesseract-OCR\tesseract.exe'
for (startX, startY, endX, endY) in boxes:
# scaling the bounding box coordinates based on the respective ratios
startX = int(startX * rW)
startY = int(startY * rH)
endX = int(endX * rW)
endY = int(endY * rH)
# applying padding in percentage
dX = int((endX - startX) * args["padding"])
dY = int((endY - startY) * args["padding"])
# apply padding to each side of the bounding box, respectively
startX = max(0, startX - dX)
startY = max(0, startY - dY)
endX = min(W, endX + (dX * 2))
endY = min(H, endY + (dY * 2))
# extract the actual padded image out of original
roi = orig[startY:endY, startX:endX]
# config stating language, LSTM model and stating all is one line of text
config = ("-l eng --oem 1 --psm 7")
# obtaining text out of image
text = filterText_video(
pytesseract.image_to_string(roi, config=config))
# text = pytesseract.image_to_string(roi, config=config)
if text and text not in predictedTexts:
# add the bounding box coordinates and text
print("Predicted Text")
print("========")
# timestamps for webcam wil be in realtime whereas for webcam will be according to webcam
if args["webcam"]:
print(text, " at time ~ ", end="")
printTime(vs.get(cv2.CAP_PROP_POS_MSEC)
) # converting millisecs into hour min secs
else:
Tt = datetime.datetime.now().strftime(
"%H:%M:%S on %d/%m/%Y")
T.append(Tt)
predictedTexts.append(text)
# draw the bounding box on the frame
cv2.rectangle(orig, (startX, startY), (endX, endY),
(0, 255, 0), 2)
os.chdir(r"C:\Users\hp\Desktop\ml\static")
rnd = time.time()
fname = str(rnd) + ".jpg"
cv2.imwrite(fname, orig)
os.chdir(r"C:\Users\hp\Desktop\ml")
vfname.append(fname)
fps.update()
# show the output frame
cv2.imshow("Text Detection", orig)
# if the `q` key was pressed, break from the loop
key = cv2.waitKey(1) & 0xFF
if key == ord("q"):
break
# stop the timer and display FPS information
fps.stop()
print("[INFO] elasped time: {:.2f}".format(fps.elapsed()))
print("[INFO] approx. FPS: {:.2f}".format(fps.fps()))
# if we are using a webcam, release the pointer
if not args["webcam"]:
vs.stop()
# otherwise, release the file pointer of webcam
else:
vs.release()
# close all windows
cv2.destroyAllWindows()
data = pd.read_csv("C:\\Users\\hp\\Desktop\\ml\\static\\" + args["excel"])
df = pd.DataFrame(data, columns=['Bib_no'])
bib_list = df.values.tolist()
print('bib_list : ', bib_list)
print('predictedTexts : ', predictedTexts)
n = len(bib_list)
pred = []
for el in predictedTexts:
sub = el.split(', ')
pred.append(sub)
print('pred:', pred)
if (len(pred) != n):
pred.append(None)
print('pred:', pred)
p = len(pred)
b = []
for i in range(n):
for j in range(p):
if (bib_list[i] == pred[j]):
df.loc[i, 'Status'] = T[j]
#df.loc[i, 'Time and Date'] = T[j]
break
else:
df.loc[i, 'Status'] = 'Not Predicted'
#df.Status.fillna("Not predicted", inplace=True)
print(df)
download_source = (
r'C:\Users\hp\Desktop\ml\static\Wexcel\output_video.xlsx')
df.to_excel(download_source)
return predictedTexts, vfname, T
mon = {'top': 0, 'left': 0, 'width': 1280, 'height': 720}
sct = None
if "screen" in url:
sct = mss.mss()
webcam=False
#webcam=True
cap = None
if sct is None:
if webcam:
#cap = WebcamVideoStream(src=""+str(videoUrl)+"").start()
cap = WebcamVideoStream(videoUrl).start()
else:
cap = cv2.VideoCapture(videoUrl)
cap.set(cv2.CAP_PROP_FRAME_WIDTH, procWidth)
cap.set(cv2.CAP_PROP_FRAME_HEIGHT, procHeight)
#cap = cv2.VideoCapture(videoUrl)
count=50
#skip=2000
skip=skipNr
SKIP_EVERY=150 #pick a frame every 5 seconds
count=1000000
#skip=0 #int(7622-5)
SKIP_EVERY=0
every=SKIP_EVERY
