I have a MacBook Air M1 2020, Ventura, 13.1 OS.
This is my code to capture a video
cap = cv2.VideoCapture(0)
width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
writer = cv2.VideoWriter('video_out.mp4', cv2.VideoWriter_fourcc('m', 'p', '4', 'v'),12, (width, height))
while True:
ret, frame = cap.read()
frame = detect_face(frame)
writer.write(frame)
cv2.imshow('Video Face Detection', frame)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
cap.release()
writer.release()
cv2.destroyAllWindows()
The detect face function.
def detect_face(img):
flag = False
face_img = img.copy()
face_rects = face_cascade.detectMultiScale(face_img)
for (x,y,w,h) in face_rects:
cv2.rectangle(face_img, (x,y), (x+w,y+h), (255,255,255), 2)
crop_img = face_img[y:y+h,x:x+w]
flag = True
crop = crop_img.copy()
crop = cv2.resize(crop,(48,48))
crop = crop.reshape(1,48,48,3)
crop = crop.astype('float64')
pred = model_pred.predict(crop)
n1 = int(np.random.randint(0,256,1))
n2 = int(np.random.randint(0,256,1))
n3 = int(np.random.randint(0,256,1))
text = exp[pred.argmax()]
cv2.putText(face_img,text,(x,y-30),fontFace = cv2.FONT_ITALIC,
fontScale = 2,color=[n1,n2,n3],thickness=7)
crop_img = cv2.resize(crop_img,(227,227))
blob = cv2.dnn.blobFromImage(crop_img, 1, (227, 227), MODEL_MEAN_VALUES, swapRB=False)
age_net.setInput(blob)
age_preds = age_net.forward()
age = age_list[age_preds[0].argmax()]
text = 'Age : ' + age
cv2.putText(face_img,text,(x-20,y+h+50),fontFace = cv2.FONT_HERSHEY_SIMPLEX,
fontScale = 1,color=[0,220,0],thickness=3)
gender_net.setInput(blob)
gender_preds = gender_net.forward()
gender = gender_list[gender_preds[0].argmax()]
text = 'Gender : '+ gender
cv2.putText(face_img,text,(x+w+2,y+int(h/2)+20),fontFace = cv2.FONT_ITALIC,
fontScale = 1,color=[0,0,255],thickness=3)
if(flag):
return face_img
else:
return face_img
I have tried so many versions but I am unable to capture a video. I always end up getting this image in a frame.
I don't understand how or why this is getting captured. Any help is appreciated.
How about first making sure the camera has captured something?
ret, frame = cap.read()
if ret:
# run the face detector
else:
break
Related
I have a logo and I am trying to put this logo into my video. When I add the logo with below codes, logo couldnt be original color, it shows transparent. But I dont want to transparency I need to put original color format.
I get this output:
This is my code:
img_path = 'ap_logo.png'
logo = cv2.imread(img_path,cv2.IMREAD_UNCHANGED)
#
#
watermark = image_resize(logo, height=300)
watermark = cv2.cvtColor(watermark, cv2.COLOR_BGR2BGRA)
watermark_h, watermark_w, watermark_c = watermark.shape
ret, frame = cap.read()
frame = cv2.flip(frame, 1)
frame = cv2.rotate(frame, cv2.ROTATE_90_CLOCKWISE)
frame_h, frame_w, frame_c = frame.shape
# # overlay with 4 channel BGR and Alpha
overlay = np.zeros((frame_h, frame_w, 4), dtype='uint8')
for i in range(0, watermark_h):
print(i)
for j in range(0, watermark_w):
if watermark[i, j][3] != 0:
h_offset = frame_h - watermark_h
w_offset = frame_w - watermark_w
overlay[h_offset + i, w_offset + j] = watermark[i, j]
while(True):
ret, frame = cap.read()
frame = cv2.flip(frame, 1)
frame = cv2.rotate(frame, cv2.ROTATE_90_CLOCKWISE)
frame = cv2.cvtColor(frame, cv2.COLOR_BGR2BGRA)
cv2.addWeighted(overlay, 0.25, frame, 1.0, 0, frame)
# Display the resulting frame
frame = cv2.cvtColor(frame, cv2.COLOR_BGRA2BGR)
if ret:
cv2.imshow('Frame', frame)
out.write(frame) # file a ilgili frame yazılıyor
if cv2.waitKey(1) & 0xFF == ord('q'):
break
#When everything done, relase the capture
cap.release()
out.release() # saved
cv2.destroyAllWindows()
I want to extract last frame of video that have been processed before with Motion History Image (MHI) and I want to save last image processed by MHI.
Here is my code:
import numpy as np
import cv2
MHI_DURATION = 35
DEFAULT_THRESHOLD = 30
def main():
live_video = False
video_src = 1
if not live_video:
video_src = "data/vid1.mp4"
cv2.namedWindow('motion-history')
cv2.namedWindow('raw')
cv2.moveWindow('raw', 200, 0)
cam = cv2.VideoCapture(video_src)
ret, frame = cam.read()
h, w = frame.shape[:2]
prev_frame = frame.copy()
motion_history = np.zeros((h, w), np.float32)
timestamp = 0
while True:
ret, frame = cam.read()
if not ret:
break
frame_diff = cv2.absdiff(frame, prev_frame)
gray_diff = cv2.cvtColor(frame_diff, cv2.COLOR_BGR2GRAY)
ret, fgmask = cv2.threshold(gray_diff, DEFAULT_THRESHOLD, 1, cv2.THRESH_BINARY)
timestamp += 1
# update motion history
cv2.motempl.updateMotionHistory(fgmask, motion_history, timestamp, MHI_DURATION)
# normalize motion history
mh = np.uint8(np.clip((motion_history - (timestamp - MHI_DURATION)) / MHI_DURATION, 0, 1) * 255)
cv2.imshow('motion-history', mh)
# cv2.imshow('raw', frame)
prev_frame = frame.copy()
if 0xFF & cv2.waitKey(5) == 27:
break
cv2.destroyAllWindows()
And, can we process faster without loop through all frame in the video?
import numpy as np
import cv2
cap = cv2.VideoCapture('car.avi')
size = (int(cap.get(cv2.cv.CV_CAP_PROP_FRAME_WIDTH)), int(cap.get(cv2.cv.CV_CAP_PROP_FRAME_HEIGHT)))
fourcc = cv2.cv.CV_FOURCC(*'AVID')
video = cv2.VideoWriter('6.avi', fourcc, 25, size)
while(1):
ret, frame = cap.read()
if not ret:
break
frame = cv2.convertScaleAbs(frame)
params = cv2.SimpleBlobDetector_Params()
params.blobColor = 0
params.filterByColor = True
params.minArea = 0
params.filterByArea = False
params.minThreshold = 120;
params.maxThreshold = 255;
detector = cv2.SimpleBlobDetector(params)
keypoints = detector.detect(frame)
im_with_keypoints = cv2.drawKeypoints(frame, keypoints, np.array([]), (0,0,255), cv2.DRAW_MATCHES_FLAGS_DRAW_RICH_KEYPOINTS)
if ret == True:
video.write(im_with_keypoints)
cv2.imshow('frame', im_with_keypoints)
else:
cap.release()
video.release()
break
k = cv2.waitKey(10) & 0xff
if k == 27:
break
I modified the code you have, and now it can run. I used python3.5.
import numpy as np
import cv2
cap = cv2.VideoCapture(r'E:/test.mp4')
size = (int(cap.get(cv2.CAP_PROP_FRAME_WIDTH)), int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT)))
fourcc = cv2.VideoWriter_fourcc(*'DIVX')
video = cv2.VideoWriter(r'E:/6.avi', fourcc, 25, size)
while(1):
ret, frame = cap.read()
if not ret:
break
frame = cv2.convertScaleAbs(frame)
params = cv2.SimpleBlobDetector_Params()
params.blobColor = 0
params.filterByColor = True
params.minArea = 0
params.filterByArea = False
params.minThreshold = 120;
params.maxThreshold = 255;
ver = (cv2.__version__).split('.')
if int(ver[0]) < 3:
detector = cv2.SimpleBlobDetector(params)
else:
detector = cv2.SimpleBlobDetector_create(params)
keypoints = detector.detect(frame)
im_with_keypoints = cv2.drawKeypoints(frame, keypoints, np.array([]), (0,0,255), cv2.DRAW_MATCHES_FLAGS_DRAW_RICH_KEYPOINTS)
if ret == True:
video.write(im_with_keypoints)
cv2.imshow('frame', im_with_keypoints)
else:
cap.release()
break
k = cv2.waitKey(10) & 0xff
if k == 27:
break
Basically, this code will detect object's motion in the scene. When the motion is detected, a red rectangle frame will track the object's motion. But, I added a new function into the code which is frame differencing. In general it is thresholding. When i run the code it says:"cv2.capture" object is not callable.
import cv2.cv as cv
class Target:
def __init__(self):
self.capture = cv.CaptureFromCAM(0)
cv.NamedWindow("Target", 1)
def run(self):
# Capture first frame to get size
ret, current_frame = self.capture()
previous_frame = current_frame
frame = cv.QueryFrame(self.capture)
frame_size = cv.GetSize(frame)
color_image = cv.CreateImage(cv.GetSize(frame), 8, 3)
grey_image = cv.CreateImage(cv.GetSize(frame), cv.IPL_DEPTH_8U, 1)
moving_average = cv.CreateImage(cv.GetSize(frame), cv.IPL_DEPTH_32F, 3)
first = True
while True:
current_frame_gray = cv2.cvtColor(current_frame,cv2.COLOR_BGR2GRAY)
previous_frame_gray = cv2.cvtColor(previous_frame, cv2.COLOR_BGR2GRAY)
frame_diff = cv2.absdiff(current_frame_gray,previous_frame_gray)
closest_to_left = cv.GetSize(frame)[0]
closest_to_right = cv.GetSize(frame)[1]
color_image = cv.QueryFrame(self.capture)
# Smooth to get rid of false positives
cv.Smooth(color_image, color_image, cv.CV_GAUSSIAN, 3, 0)
if first:
difference = cv.CloneImage(color_image)
temp = cv.CloneImage(color_image)
cv.ConvertScale(color_image, moving_average, 1.0, 0.0)
first = False
else:
cv.RunningAvg(color_image, moving_average, 0.020, None)
# Convert the scale of the moving average.
cv.ConvertScale(moving_average, temp, 1.0, 0.0)
# Minus the current frame from the moving average.
cv.AbsDiff(color_image, temp, difference)
# Convert the image to grayscale.
cv.CvtColor(difference, grey_image, cv.CV_RGB2GRAY)
# Convert the image to black and white.
cv.Threshold(grey_image, grey_image, 70, 255, cv.CV_THRESH_BINARY)
# Dilate and erode to get people blobs
cv.Dilate(grey_image, grey_image, None, 18)
cv.Erode(grey_image, grey_image, None, 10)
storage = cv.CreateMemStorage(0)
contour = cv.FindContours(grey_image, storage, cv.CV_RETR_CCOMP, cv.CV_CHAIN_APPROX_SIMPLE)
points = []
while contour:
bound_rect = cv.BoundingRect(list(contour))
contour = contour.h_next()
pt1 = (bound_rect[0], bound_rect[1])
pt2 = (bound_rect[0] + bound_rect[2], bound_rect[1] + bound_rect[3])
points.append(pt1)
points.append(pt2)
cv.Rectangle(color_image, pt1, pt2, cv.CV_RGB(255,0,0), 1)
cv.ShowImage("Target", color_image)
cv.ShowImage("frame_diff", frame_diff)
# Listen for ESC key
c = cv.WaitKey(7) % 0x100
if c == 27:
break
if __name__=="__main__":
t = Target()
t.run()
This is the original frame differencing code that I wanted transfer from:
import cv2
cap = cv2.VideoCapture(0)
ret, current_frame = cap.read()
previous_frame = current_frame
while(cap.isOpened()):
current_frame_gray = cv2.cvtColor(current_frame, cv2.COLOR_BGR2GRAY)
previous_frame_gray = cv2.cvtColor(previous_frame, cv2.COLOR_BGR2GRAY)
frame_diff = cv2.absdiff(current_frame_gray,previous_frame_gray)
cv2.imshow('frame diff ',frame_diff)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
previous_frame = current_frame.copy()
ret, current_frame = cap.read()
cap.release()
cv2.destroyAllWindows()
You can't call the capture object, but you can read from it by calling its read method. Changing your frame return code to:
ret, current_frame = self.capture.read()
Should fix this.
Edit:
The two lines:
ret, current_frame = self.capture.read()
previous_frame = current_frame
Will work for VideoCapture but not CaptureFromCAM. You apparently don't need them since you can take a frame by doing:
frame = cv.QueryFrame(self.capture)
I am trying to understand how to work with ROI, but the following code is not working as expected.
I have a white ball walking through a black background, and I want to detect and keep tracking of this ball. Unfortunately, the square does not move.I think the problem is the color array, but after trying many values, I still haven't found the correct solution.
import numpy as np
import cv2
cap = cv2.VideoCapture('oieee.avi')
ret,frame = cap.read()
r,h,c,w = 485,50,890,50
track_window = (c,r,w,h)
roi = frame[r:r+h, c:c+w]
hsv_roi = cv2.cvtColor(frame, cv2.COLOR_BGR2HSV)
mask = cv2.inRange(hsv_roi, np.array((0., 0., 240.)), np.array((255.,15.,255.)))
roi_hist = cv2.calcHist([hsv_roi],[0],mask,[180],[0,180])
cv2.normalize(roi_hist,roi_hist,0,255,cv2.NORM_MINMAX)
term_crit = ( cv2.TERM_CRITERIA_EPS | cv2.TERM_CRITERIA_COUNT, 10, 1 )
while(1):
ret ,frame = cap.read()
if ret == True:
hsv = cv2.cvtColor(frame, cv2.COLOR_BGR2HSV)
dst = cv2.calcBackProject([hsv],[0],roi_hist,[0,180],1)
ret, track_window = cv2.meanShift(dst, track_window, term_crit)
x,y,w,h = track_window
cv2.rectangle(frame, (x,y), (x+w,y+h), 255,2)
img2 = cv2.rectangle(frame, (x,y), (x+w,y+h), 255,2)
cv2.imshow('img2',frame)
k = cv2.waitKey(60) & 0xff
if k == 27:
break
else:
cv2.imwrite(chr(k)+".jpg",img2)
else:
break
cv2.destroyAllWindows()
cap.release()
A print screen of my video http://imgur.com/VG3TgFF