Attribute of Video and reference pixel value

Attribute of Video and reference pixel value

1. Open CV express the video data to use numpy.ndarray

import cv2 img1 = cv2.imread(‘cat.bmp’,cv2.IMREAD_GRAYSCALE) img2 = cv2.imread(‘cat.bmp’,cv2.IMREAD_COLOR)

• ndim : the number of dimension. = len(img.shape)
• shape : the size of each dimension. * Grayscale = (h,w), Colorscale = (h,w,3)
• size : Total number of elements
• dytp : Data type of elements. ex) Video data is uint8

2. OpenCV data type and Numpy data type

cv.CV_8U	numpy.unint8	8-bit unsigned integer
cv.CV_16S	numpy.int16	16-bit signed integer
cv.CV_32F	numpy.float32	32-bit floating point

• Grayscale Video : cv2.CV_8UC1 -> numpy.uint8, shape = (h,w)
• Colorscale Video : cv2.CV_8UC1 -> numpy.uint8, shape = (h,w,3)

3. Mask Operation and ROI

• ROI
  • Region of Interest
  • Specific region that could special operation in the video.
• Mask Operation
  • OpenCV support ROI operaion to some of function when it do the mask Video should be converted.
  • Generally, Mask video is used through the binary image ,which is composed of 0 or 255.
• Pixel value copy function supporting mask operation.

cv2.copyTo(src,maks,dst=None) -> dst

• src : input video
• mask : mask video
• sdt : output layer