def images_to_video(vout, image_dir, clear_images=True):
image_list = glob.glob(f'{timelapse_img_path}/{i}.jpg')
sorted_images = sorted(image_list, key=os.path.getmtime)
if file in sorted_images:
image_frame = cv2.read(file)
vout.write(image_frame)
if clear_images:
for file in image_list:
os.remove(file)
def load_imgs():
global imgs
global wheels
for p in purposes:
for epoch_id in epochs[p]:
print(
'processing and loading "{}" epoch {} into memory, current num of imgs is {}...'
.format(p, epoch_id, len(imgs[p])))
# vid_path = cm.jn(data_dir, 'epoch{:0>2}_front.mkv'.format(epoch_id))
data_dir = os.path.abspath('training_images{}'.format(epoch_id))
assert os.path.isdir(data_dir)
print("upload images from:", data_dir)
# csv_path = cm.jn(data_dir, 'epoch{:0>2}_steering.csv'.format(epoch_id))
csv_path = os.path.abspath('out-key{}.csv'.format(epoch_id))
assert os.path.isfile(csv_path)
print("upload ground truth from:", csv_path)
rows = cm.fetch_csv_data(csv_path)
print(len(rows), frame_count)
assert frame_count == len(rows)
yy = [[float(row['wheel'])] for row in rows]
for image in data_dir:
img = cv2.read(image)
img = preprocess.preprocess(img)
imgs[p].append(img)
wheels[p].extend(yy)
assert len(imgs[p]) == len(wheels[p])
cap.release()
import cv2
import numpy as np
import math
img = cv2.read("image.jpg") # take image as input
gray = cv2.cvtColor(img,COLOR_BGR2GRAY) # Convert image to grayscale
edges = cv2.Canny(gray,50,150, apertureSize = 3) #Apply Canny edge detector method on the grayscale image
lines = cv2.HoughLines(edges,1,np.pi/180,200)
print("Dip of the lines in the image")
sum=0
for i in range(0,len(lines)):
for rho,thetha in lines[i]:
sum+= np.pi/2- thetha
print(np.pi/2-thetha)
print("Average value of dip is:" sum/len(lines))
num = len(lines)
x,y,dim = img.shape
dip = sum/num # average dip value in radian
m = num/(num+1)
'''
****Persistence length****
Persistence of discontinuities is one of the most
important rock mass parameters because it
defines, together with spacing, the size of blocks
that can slide from the face
def main():
"""main function"""
def process_frame_pair(frames, color) -> bool:
left = frames[0]
right = frames[1]
# make sure images are the same size
assert left.shape == right.shape
# get depth map with rectified images
depth_map = get_depth_map(left, right, stereo)
# plot image with depth
if POINTCLOUD:
plot_image(pc, np.rot90(colorized, -1), np.rot90(depth_map, -1) * (1.0 / POINT_CLOUD_SCALE), POINT_CLOUD_SCALE, step=1)
grad_x, grad_y = np.gradient(cv2.GaussianBlur(depth_map, (25, 25), 25.0) * 33.3, 1)
direction, points = get_direction(depth_map, grad_x, grad_y, 250, 1.0)
debug = np.zeros(depth_map.shape)
for point in points:
cv2.circle(debug, (int(point[0]), int(point[1])), 1, (255, 255, 255), 1)
if cv2.waitKey(1) & 0xFF != ord('q') and (not POINTCLOUD or pc.render()):
cv2.imshow('left', rectified_left)
cv2.imshow('depth', depth_map)
cv2.imshow('debug', debug)
return False
# object to compute disparity map
stereo: cv2.StereoBM = cv2.StereoBM_create(constants.NUM_DISPARITIES, constants.BLOCK_SIZE)
# custom point cloud renderer
if POINTCLOUD:
pc: point_cloud.PointCloud = point_cloud.PointCloud("3D Representation", 2.0)
# load saved calibration
calibration = np.load("test_calibration.npz")
img_size = calibration["img_size"]
left_map_x = calibration["left_map_x"]
left_map_y = calibration["left_map_y"]
right_map_x = calibration["right_map_x"]
right_map_y = calibration["right_map_y"]
right_map_y = calibration["right_map_y"]
# noinspection PyUnboundLocalVariable
frames = 0
durations = []
frames_per_second = []
while True: # not cameras or (cameras and capture_l.isOpened() and cameras and capture_r.isOpened()):
pre = time.time()
# save colorized image to render point cloud
colorized = read('tests/test_left.png') / 255.0
img_left, img_right = read('tests/test_left.png', cv2.CV_8UC1), read('tests/test_right.png', cv2.CV_8UC1)
# remap images to align normals of image planes
rectified_left = cv2.remap(img_left, left_map_x, left_map_y, REMAP_INTERPOLATION)
rectified_right = cv2.remap(img_right, right_map_x, right_map_y, REMAP_INTERPOLATION)
if DEBUG:
rectified_left = img_left
rectified_right = img_right
if process_frame_pair((rectified_left, rectified_right), colorized):
break
durations.append(dur := round(time.time() - pre, 3))
frames_per_second.append(fps := round(1.0 / dur, 3))
print(
"\rProcessed Frame: {frames} (took {dur:<5} ms, avg: {avg:<7} | fps: {fps:<7}, avg fps: {avg_fps:<7})".format(
frames=(frames := frames + 1),
dur=dur,
avg=round(float(np.mean(durations)), 3),
fps=fps,
avg_fps=round(float(np.mean([1.0 / x for x in durations])), 3)),
end='')
cv2.destroyAllWindows()
import cv2
import time
cam=cv2.VideoCapture(0)
start=time.time()
end=time.time()
while True:
x,red=cv2.read()
if end-start==5:
flipped=cam.flip(red,-1)
cv2.imshow('win',flipped)
else:
cv2.imshow('win',red)
if cv2.waitKey(1)&0xFF==ord('q'):
break
import cv2 as cv
import numpy as np
cap=cv.VideoCapture()
ret,first_frame=cv.read()
prev_gray=cv.cvtColor(first_frame,cv.COLOR_BGR2GRAY)
mask=np.zeros_like(frame)
mask[..., 1]=255
while (cap.isOpened()):
ret, frame = cap.read()
cv.imshow('input',frame)
gray = cv.cvtColor(frame, cv2.COLOR_BGR2GRAY)
flow= cv.calcOpticalFlowFarneback(prev_gray,
gray,
None,
0.5,3,15,3,5,1.2,0)
magn, angle = cv.cartToPolar((flow[...,0],
flow[...,1]))
mask[...,2] = cv.normalize(magn,
None,
0,
255,
cv.NORM_MINMAX)
rgb = cv.cvtColor(mask, cv.COLOR_HSV2RGB)
cv.imshow("Dense Optical Flow",rgb)
prev_gray = gray
if cv.waitKey(30) & 0xFF == ord("q"):
break
cap.release()
cv.destroyAllWindows()
import cv2
cap = cv2.VideoCapture(0)
# Create the haar cascade
faceCascade = cv2.CascadeClassifier("../model/haarcascade_frontalface_alt.xml")
img_path = ''
while (True):
# Capture frame-by-frame
ret, frame = cap.read()
cv2.read()
# Our operations on the frame come here
gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
# Detect faces in the image
faces = faceCascade.detectMultiScale(gray,
scaleFactor=1.1,
minNeighbors=5,
minSize=(30, 30)
#flags = cv2.CV_HAAR_SCALE_IMAGE
)
print("Found {0} faces!".format(len(faces)))
# Draw a rectangle around the faces
for (x, y, w, h) in faces:
cv2.rectangle(frame, (x, y), (x + w, y + h), (0, 255, 0), 2)
# Display the resulting frame
# pre requisits : 1_basics.py
import cv2
img=cv2.read("imageProcessing.jpg")
height,width=img.shape[0:2]
# ------ resize Image ------
# resize(src,dsize,fx,fy)
# src = source of the image
# dsize = desired size of the output image
# fx = scalling factor along x-axis
# fy = scalling factor along y-axis
# method-1
resizeImg=cv2.resize(img,(0,0),fx=1.75,fy=1.75)
# method-2
resizeImg1=cv2.resize(img,(550,350))
cv2.imshow('resized image',resizeImg)
cv2.imshow('resized image 1',resizeImg1)
cv2.waitKey(0)
# dimensions to crop the image frame
y1 = 0
x1 = 0
y2 = 500
x2 = 500
# read video feed from any USB camera
feed = cv2.VideoCapture(0) # 0 denotes in-built Webcam
if not feed.isOpened():
raise IOError("Webcam cannot be opened")
while True:
# Extract frames
ret, frame = cv2.read()
# show run-time video feed on console
cv2.imshow("input", frame)
# crop image to user-defined size
cropped_image = frame[y1:y2, x1:x2]
# write frame to the local system
cv2.imwrite(output + "/%#05d.jpg" % (count + 1), cropped_image)
# get path of the saved frame
path = output + "/%#05d.jpg" % (count + 1)
count = count + 1
def main(args):
print('################################################################################')
my_style_data=[]
with tf.Graph().as_default():
with tf.Session() as sess:
# Setup models
mtcnn = detect_and_align.create_mtcnn(sess, None) #It calls create_mtcnn function from the detect_and_align file
load_model(args.model) #IT loads the facenet 20170512-110547.pb pre-trained model
images_placeholder = tf.get_default_graph().get_tensor_by_name("input:0")
embeddings = tf.get_default_graph().get_tensor_by_name("embeddings:0")
phase_train_placeholder = tf.get_default_graph().get_tensor_by_name("phase_train:0")
# Load anchor IDs
id_data = IdData(args.id_folder[0], mtcnn, sess, embeddings, images_placeholder, phase_train_placeholder, args.threshold)
#url ='rtsp://192.168.137.135:4747/video'
#cap = cv2.VideoCapture(0)
connection = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
connection.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
connection.settimeout(TIMEOUT_SOCKET)
connection.connect((IP_SERVER, PORT_SERVER))
while True:
try:
fileDescriptor = connection.makefile(mode='rb')
result = fileDescriptor.readline()
fileDescriptor.close()
result = base64.b64decode(result)
frame = np.fromstring(result, dtype=np.uint8)
frame_matrix = np.array(frame)
frame_matrix = np.reshape(frame_matrix, (IMAGE_HEIGHT, IMAGE_WIDTH,COLOR_PIXEL))
xyz=cv2.imshow('Window title', frame_matrix)
cv2.read(xyz)
print("cccc")
frame_height = frame_matrix.get(cv2.CAP_PROP_FRAME_HEIGHT)
show_landmarks = True
show_bb = True
show_id = True
show_fps = False
show_bb1 = True
while(True):
start = time.time()
v_offset = 50
time.sleep(0.0001)
_, frame = frame_matrix.read()
frame1=frame
# Locate faces and landmarks in frame
face_patches, padded_bounding_boxes, landmarks = detect_and_align.detect_faces(frame, mtcnn)
if len(face_patches) > 0:
face_patches = np.stack(face_patches)
feed_dict = {images_placeholder: face_patches, phase_train_placeholder: False}
embs = sess.run(embeddings, feed_dict=feed_dict)
print('Matches in frame:')
matching_ids, matching_distances = id_data.find_matching_ids(embs)
for bb, landmark, matching_id, dist in zip(padded_bounding_boxes, landmarks, matching_ids, matching_distances):
if matching_id is None:
matching_id = 'Unknown'
print('Unknown! Couldn\'t fint match.')
else:
#if(int(bb[0])<=170 and int(bb[3])<=357 and int(bb[2])<=437 and int(bb[1])<=164):
print('Hi akki %s! Distance: %1.4f' % (matching_id, dist))
now=datetime.now()
#csvData = [matching_id, dist,now.strftime("%x %I:%M:%S %p")]
'''with open('C:/myproject/Phase2-Copy_FaceRecognition-master/Student4.csv', 'a') as csvFile:
writer = csv.writer(csvFile)
writer.writerow(csvData)
csvFile.close()'''
if show_id:
font = cv2.FONT_HERSHEY_SIMPLEX
print("bb_Frame",bb)
if(210<int(bb[0])<350 and 150<int(bb[1])<250 and 300<int(bb[2])<460 and 310<int(bb[3])<450):
welcome=" Welcome to Infogen labs"
cv2.putText(frame,matching_id+ welcome, (0, 50), font, 1, (0,0,255), 2, cv2.LINE_AA)
cv2.putText(frame,matching_id+now.strftime(" %I:%M%p"), (bb[0], bb[3]), font, 1, (0,0,255), 2, cv2.LINE_AA)
#csvData = [matching_id, dist,now.strftime("%x %I:%M:%S %p")]
def previous_and_next(some_iterable):
prevs, items, nexts = tee(some_iterable, 3)
prevs = chain([None], prevs)
nexts = chain(islice(nexts, 1, None), [None])
return zip(prevs, items, nexts)
csvData = [matching_id, dist,now.strftime("%x %I:%M:%S %p")]
with open('C:/myproject/1Copy_FaceRecognition-master - Copy/Student5.csv', 'a') as csvFile1:
writer = csv.writer(csvFile1)
writer.writerow(csvData)
csvFile1.close()
my_style_data.append(csvData)
print('******************************************************************************************')
print(my_style_data)
print('******************************************************************************************')
if len(my_style_data) >=7:
for prevs, item, nxt in previous_and_next(my_style_data):
try:
if prevs[0] != item[0] or item[0] != nxt[0]:
with open('C:/myproject/1Copy_FaceRecognition-master - Copy/Employee6.csv', 'a') as csvFile:
writer = csv.writer(csvFile)
writer.writerow(item)
csvFile.close()
except: pass
del(my_style_data[0:7])
'''if(matching_id=="Akshay"):
with open('C:/myproject/Phase2-Copy_FaceRecognition-master/Akshay.csv', 'a') as csvFile:
writer = csv.writer(csvFile)
writer.writerow(csvData)
csvFile.close()
if(matching_id=="Ajinkya"):
with open('C:/myproject/Phase2-Copy_FaceRecognition-master/Ajinkya.csv', 'a') as csvFile:
writer = csv.writer(csvFile)
writer.writerow(csvData)
csvFile.close()'''
if show_bb:
cv2.rectangle(frame, (bb[0], bb[1]), (bb[2], bb[3]), (255, 0, 0), 1)
if show_bb1:
cv2.rectangle(frame1, (261,174),(457,380), (255,0,255),2)
if show_landmarks:
for j in range(5):
size = 1
x=(int(landmark[j]))
top_left = (int(landmark[j]) - size, int(landmark[j + 5]) - size)
bottom_right = (int(landmark[j]) + size, int(landmark[j + 5]) + size)
cv2.rectangle(frame, top_left, bottom_right, (255, 0, 255), 2)
else:
print('Couldn\'t find a face')
cv2.imshow('frame', frame)
key = cv2.waitKey(100)
if key == ord('q'):
break
elif key == ord('l'):
show_landmarks = not show_landmarks
elif key == ord('b'):
show_bb = not show_bb
elif key == ord('i'):
show_id = not show_id
elif key == ord('f'):
show_fps = not show_fps
frame_matrix.release()
cv2.destroyAllWindows()
class Face_recog:
""" this is for screen"""
def __init__(self,root):
self.root=root
self.root.configure(bg='black')
self.root.geometry("1500x770+0+0")
self.root.title("Attendence management system")
#title label
title_lbl=Label(self.root, text="Face Detector",
font=("times new roman",40,"bold","italic"),
bg="white",
fg="purple"
)
title_lbl.place(x=0,y=0,width=1500,height=45)
img=Image.open(r"C:\Users\ANCHAL\Desktop\dice_scrolling\face_detection_img\img1.jpg")
img=img.resize((500,150),Image.ANTIALIAS) # Antialas converts high level image into low lwvwl img
self.photoimg=ImageTk.PhotoImage(img)
f_lbl=Label(self.root, image=self.photoimg)
f_lbl.place(x=0,y=45,width=500, height=150)
img1=Image.open(r"C:\Users\ANCHAL\Desktop\dice_scrolling\face_detection_img\images2.jpg")
img1=img1.resize((500,150),Image.ANTIALIAS) # Antialas converts high level image into low lwvwl img
self.photoimg1=ImageTk.PhotoImage(img1)
f_lbl=Label(self.root, image=self.photoimg1)
f_lbl.place(x=500,y=45,width=500, height=150)
img2=Image.open(r"C:\Users\ANCHAL\Desktop\dice_scrolling\face_detection_img\img1.jpg")
img2=img2.resize((500,150),Image.ANTIALIAS) # Antialas converts high level image into low lwvwl img
self.photoimg2=ImageTk.PhotoImage(img2)
f_lbl=Label(self.root, image=self.photoimg2)
f_lbl.place(x=1000,y=45,width=500, height=150)
img6=Image.open(r"C:\Users\ANCHAL\Desktop\face_recognition\face_detection_img\circle.jpg")
img6=img6.resize((500,700),Image.ANTIALIAS) # Antialas converts high level image into low lwvwl img
self.photoimg6=ImageTk.PhotoImage(img6)
f_lbl=Label(self.root, image=self.photoimg6)
f_lbl.place(x=500,y=300,width=500, height=300)
detect_btn=Button(self.root,text="Detect Face",command=self.face_detect,font=("times new roman",25,"bold"),bg="black",fg="white",width=10)
detect_btn.place(x=640,y=430)
# face_recognition
def face_detect(self):
faceDetect=cv2.CascadeClassifier('haarcascade_frontalface_default.xml')
cam=cv2.VideoCapture(0)
rec=cv2.createLBPHFaceRecognizer()
rec.load("classifier.xml")
font=cv2.cv.InitFont(cv2.cv.CV_FONT_HERSHEY_COMPLEX,0.4,1,0,1)
def getProfile(id):
conn=mysql.connector.connect(host="localhost",username="******",password="******",database="face_recognition")
my_cursur=conn.cursor()
while(True):
ret,img=cv2.read()
gray=cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
faces=faceDetect.detectMultiScale(gray,1.3,5)
for(x,y,w,h) in faces:
cv2.rectangle(img,(x,y),(x+w,y+h),(0,255,0),2)
id,conf=rec.predict(gray[y:y+h,x:x+w])
profile=getProfile(id)
if(profile!=None):
cv2.cv.PutText(cv2.cv.fromarray(img),"Name : "+str(profile[1]),(x,y+h+20),font,(0,255,0))
cv2.cv.PutText(cv2.cv.fromarray(img),"Age : "+str(profile[2]),(x,y+h+45),font,(0,255,0))
cv2.cv.PutText(cv2.cv.fromarray(img),"Gender : "+str(profile[3]),(x,y+h+70),font,(0,255,0))
cv2.cv.PutText(cv2.cv.fromarray(img),"Criminal Records : "+str(profile[4]),(x,y+h+95),font,(0,0,255))
cv2.imshow("Face",img)
if(cv2.waitKey(1)==ord('q')):
break
cam.release()
cv2.destroyAllWindows()
def convert(image):
i = cv2.read(image_path)
gray_image = cv2.cvtColor(i, cv2.COLOR_BGR2GRAY)
threshold_img = cv2.threshold(gray_image, 0, 255,
cv2.THRESH_BINARY | cv2.THRESH_OTSU)[1]
kernel = np.ones((1, 1), np.uint8)
threshold_img = cv2.dilate(threshold_img, kernel, iterations=1)
threshold_img = cv2.erode(threshold_img, kernel, iterations=1)
#configuring parameters for tesseract
from pytesseract import Output
custom_config = r'--oem 3 --psm 6'
# now feeding image to tesseract
details = pytesseract.image_to_data(threshold_img,
output_type=Output.DICT,
config=custom_config)
print(details.keys())
from pytesseract import Output
custom_config = r'--oem 3 --psm 6'
total_boxes = len(details['text'])
for sequence_number in range(total_boxes):
if int(details['conf'][sequence_number]) > 5:
(x, y, w, h) = (details['left'][sequence_number],
details['top'][sequence_number],
details['width'][sequence_number],
details['height'][sequence_number])
threshold_img = cv2.rectangle(threshold_img, (x, y),
(x + w, y + h), (0, 255, 0), 2)
# display image
# now feeding image to tesseract
details = pytesseract.image_to_data(threshold_img,
output_type=Output.DICT,
config=custom_config)
parse_text = []
word_list = []
last_word = ''
for word in details['text']:
if word != '':
word_list.append(word)
last_word = word
if (last_word != '' and word == '') or (word == details['text'][-1]):
parse_text.append(word_list)
word_list = []
import csv
# saving the extracted text output to a txt file
with open('result.txt', 'w', newline="") as file:
csv.writer(file, delimiter=" ").writerows(parse_text)
import pandas as pd
# reading the txt file into a dataframe to convert to csv file
df = pd.read_csv("result.txt", delimiter='\t')
df.to_csv('result.csv')
def create_pred_movie(conf, predList, moviename, outmovie, outtype, maxframes=-1):
predLocs, predscores, predmaxscores = predList
# assert false, 'stop here'
tdir = tempfile.mkdtemp()
cap = cv2.VideoCapture(moviename)
nframes = int(cap.get(cvc.FRAME_COUNT))
if maxframes > 0:
nframes = maxframes
cmap = cm.get_cmap('jet')
rgba = cmap(np.linspace(0, 1, conf.n_classes))
fig = mpl.figure.Figure(figsize=(9, 4))
canvas = FigureCanvasAgg(fig)
if conf.adjustContrast:
clahe = cv2.createCLAHE(clipLimit=2.0, tileGridSize=conf.clahegridsize)
else:
clahe = None
for curl in range(nframes):
framein = cv2.read()
framein = crop_images(framein, conf)
if framein.shape[2] > 1:
framein = framein[..., 0]
if conf.adjustContrast:
framein = clahe.apply(framein)
fig.clf()
ax1 = fig.add_subplot(1, 2, 1)
ax1.imshow(framein, cmap=cm.gray)
ax1.scatter(predLocs[curl, :, 0, 0], predLocs[curl, :, 0, 1], # hold=True,
c=cm.hsv(np.linspace(0, 1 - old_div(1., conf.n_classes), conf.n_classes)),
s=np.clip(predmaxscores[curl, :, 0] * 100, 20, 40),
linewidths=0, edgecolors='face')
ax1.axis('off')
ax2 = fig.add_subplot(1, 2, 2)
if outtype == 1:
curpreds = predscores[curl, :, :, :, 0]
elif outtype == 2:
curpreds = predscores[curl, :, :, :, 0] * 2 - 1
rgbim = create_pred_image(curpreds, conf.n_classes)
ax2.imshow(rgbim)
ax2.axis('off')
fname = "test_{:06d}.png".format(curl)
# to printout without X.
# From: http://www.dalkescientific.com/writings/diary/archive/2005/04/23/matplotlib_without_gui.html
# The size * the dpi gives the final image size
# a4"x4" image * 80 dpi ==> 320x320 pixel image
canvas.print_figure(os.path.join(tdir, fname), dpi=160)
# below is the easy way.
# plt.savefig(os.path.join(tdir,fname))
tfilestr = os.path.join(tdir, 'test_*.png')
mencoder_cmd = "mencoder mf://" + tfilestr + " -frames " + "{:d}".format(
nframes) + " -mf type=png:fps=15 -o " + outmovie + " -ovc lavc -lavcopts vcodec=mpeg4:vbitrate=2000000"
os.system(mencoder_cmd)
cap.release()
import numpy as np
import cv2
from matplotlib import pyplot as plt
img = cv2.read('mylogo.png', 0)
plt.imshow(img, cmap='gray', interpolation="bicubic")
plt.xticks([]), plt.yticks([])
plt.show()
def get_item(self, index: int):
# -> Tuple(np.ndarray, np.ndarray)
img = cv2.read(self.dataset[index])
target = self.dataset[index].split('_')
target = self._category_to_class_index(target)
return np.array(img), np.array([target])
import cv2
import numpy as np
face_cascade = cv2.CascadeClassifier('C:\opencv\sources\data\haarcascades_cuda\haarcascade_frontalface_alt.xml')
face_mask = cv2.imread('C:\Users\hp pc\Google Drive\opencv-practice\python_codes\images\\find.jpg')
h_mask, w_mask = face_mask.shape[:2]
if face_cascade.empty():
raise IOERROR('Unable to load the face cascade classifier xml file')
cap = cv2.VideoCapture(0)
scaling_factor = 0.5
while true:
ret, frame = cv2.read()
frame = cv2.resize(frame, None, fx = 0.5, fy = 0.5, Interpolation = cv2.INTER_AREA)
gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
face_rects = face_cascade.detectMultiScale(gray, 1.3, 2)
for (x, y, w, h) in face_rects:
if h > 0 and w > 0:
h, w = int(1.4*h), int(1*w)
##
# extract the region of interest(roi) from the image
frame_roi = frame[y:y+h, x:x+w]
face_mask_small = cv2.resize(face_mask, (w,h), interpolation = cv2.INTER_AREA)
# convert color image to grayscale and threshold it
os.makedirs(new_path)
# Save images to a folder
def save_data(img, steering):
global img_list, steering_list
now = datetime.now()
timestamp = str(datetime.timestamp(now)).replace('.', '')
filename = os.path.join(new_path, f'Image_{timestamp}.jpg')
cv2.imwrite(filename, img)
img_list.append(filename)
steering_list.append(steering)
# Save the log file when the session ends
def save_log():
global img_list, steering_list
raw_data = {'Image': img_list, 'Steering': steering_list}
df = pd.DataFrame(raw_data)
new_path = os.path.join(current_dir, f'log_{str(folder_count)}.csv')
df.to_csv(new_path, index=False, header=False)
print('Log saved')
print('Total Images: ', len(img_list))
if __name__ == '__main__':
cap = cv2.VideoCapture(0)
for x in range(10):
_, img = cv2.read()
save_data(img, 0.2)
cv2.waitKey(1)
cv2.imshow('Img', img)
save_log()
import cv2
import os
for user in directory:
user = []
for img in user:
cv2.read(img)
img_name = os.img_name
name, likes, comments, date, time = img_name[], img_name[], img_name[]
cap.set(cv2.CAP_PROP_FRAME_WIDTH, INPUT_SIZE[0])
cap.set(cv2.CAP_PROP_FRAME_HEIGHT, INPUT_SIZE[1])
last_capture_time = datetime.now().timestamp()
while True:
if datetime.now().timestamp() < last_capture_time + 1000 / float(options.fps):
time.sleep(0.01)
continue
try:
has_error = False
_, frame = cv2.read()
# Preprocessing
mean_shifted_image = cv2.pyrMeanShiftFiltering(frame, FILTER_SIZE, 32)
img = cv2.cvtColor(mean_shifted_image, cv2.COLOR_BGR2GRAY)
# Gaussian blur and thresholding
blurred_image = cv2.GaussianBlur(img, (FILTER_SIZE, FILTER_SIZE), 6)
thresh, binary_image = cv2.threshold(blurred_image, 0, 255, cv2.THRESH_BINARY | cv2.THRESH_OTSU)
test_img = binary_image
# Repeatedly detect and remove blobs