所以我在下面有這段代碼，我希望它能夠根據回圈i值實時將框架的寬度大小從 1 更改為 1000，以便將其可視化，當我執行代碼時，我期望它會更改視窗運行時的寬度大小。
import cv2
import numpy as np
vid = cv2.VideoCapture("C:\\users\\USER\\downloads\\man.mp4")
while True:
ret,frame = vid.read()
for i in range(1,1000):
frame = cv2.resize(frame,(i,450))
size = 16
# Create motion blur kernel
kernel_motion_blur = np.zeros((size,size))
kernel_motion_blur[int((size-1)/2), :] = np.ones(size)
kernel_motion_blur = kernel_motion_blur / size
# Apply motion kernel motion blur
result = cv2.filter2D(frame, -1, kernel_motion_blur)
cv2.imshow('Motion Blur Applied',result)
cv2.imshow('Original',frame)
if cv2.waitKey(1) == ord('q'):
break
cv2.destroyAllWindows()
vid.release()
它可以作業，但問題是當我執行代碼時，視頻幀會出現故障，并且視頻幀沒有顯示實際視頻中包含的正確影像。那么我該如何解決呢？
之前和之后的視頻
uj5u.com熱心網友回復：
我不確定我是否完全理解這個問題，或者為什么你在 while 回圈中有一個 for 回圈，但我認為我能夠達到你想要的效果。看一看：
import cv2
import numpy as np
# control how fast the image slides left to right
step_size = 10
vid = cv2.VideoCapture("videos\demo.mp4")
while True:
ret,frame = vid.read()
if ret == False:
break
width = frame.shape[1]
for n in range(0, width, step_size):
frame_to_show = frame[:,:n step_size]
size = 16
# Create motion blur kernel
kernel_motion_blur = np.zeros((size,size))
kernel_motion_blur[int((size-1)/2), :] = np.ones(size)
kernel_motion_blur = kernel_motion_blur / size
# Apply motion kernel motion blur
result = cv2.filter2D(frame_to_show, -1, kernel_motion_blur)
cv2.imshow('Motion Blur Applied',result)
cv2.imshow('Original',frame_to_show)
key = cv2.waitKey(1)
if key == ord('q'):
break
cv2.destroyAllWindows()
vid.release()
或者這可能是您正在尋找的更多內容：
import cv2
import numpy as np
# control how fast the image slides left to right
step_size = 10
vid = cv2.VideoCapture("videos\demo.mp4")
end = 0
while True:
ret,frame = vid.read()
if ret == False:
break
width = frame.shape[1]
if end > width:
end = 0
end = step_size
frame_to_show = frame[:,:end]
size = 16
# Create motion blur kernel
kernel_motion_blur = np.zeros((size,size))
kernel_motion_blur[int((size-1)/2), :] = np.ones(size)
kernel_motion_blur = kernel_motion_blur / size
# Apply motion kernel motion blur
result = cv2.filter2D(frame_to_show, -1, kernel_motion_blur)
cv2.imshow('Motion Blur Applied',result)
cv2.imshow('Original',frame_to_show)
key = cv2.waitKey(1)
if key == ord('q'):
break
cv2.destroyAllWindows()
vid.release()
轉載請註明出處，本文鏈接：https://www.uj5u.com/yidong/482838.html
上一篇：Rggplotscale_y_log10，從對數刻度中洗掉10^(-0.5)
下一篇：在gitfetch上游排除檔案夾