python language

[shiwen1997]

thank you for your excellent code,
could you please provide a new repositories written by python? Because the c++ opencv environment is too compicate.....
best wishes

[bayramf]

Take a look at here: https://github.com/qijiezhao/py-denseflow

**Edit
I see that there are several algorithms to create the optical flow. The first link uses the TVL1 algorithm. I have put together another script with the Gunnar-Farneback algorithm (with the help of: https://stackoverflow.com/questions/27644388/optical-flow-using-opencv-horizontal-and-vertical-components)

import cv2
import numpy as np
import os

# Path with image files
path = "dataset/pedestrian/"
list_names = os.listdir(path)

# Creating folder structure for rgb, x-flow and y-flow images
rgb_folder = os.mkdir(path + "rgb")
x_folder = os.mkdir(path + "x")
y_folder = os.mkdir(path + "y")

# Read in the first frame
frame1 = cv2.imread(path + list_names[0])
prvs = cv2.cvtColor(frame1,cv2.COLOR_BGR2GRAY)

# as the frame 
mask = np.zeros_like(frame1) 

# Sets image saturation to maximum 
mask[..., 1] = 255

# Set counter to read the second frame at the start
counter = 1

# Set counter for writing image files
i = 0

print("Start processing...")
# Until we reach the end of the list...
while counter < len(list_names):
    
    # Read the next frame in
    frame2 = cv2.imread(path + list_names[counter])
    next_ = cv2.cvtColor(frame2,cv2.COLOR_BGR2GRAY)

    # Calculate optical flow between the two frames
    flow = cv2.calcOpticalFlowFarneback(prvs, next_, None, 0.5, 3, 15, 3, 5, 1.2, 0)

    # Computes the magnitude and angle of the 2D vectors 
    magnitude, angle = cv2.cartToPolar(flow[..., 0], flow[..., 1]) 
      
    # Sets image hue according to the optical flow  
    # direction 
    mask[..., 0] = angle * 180 / np.pi / 2
      
    # Sets image value according to the optical flow 
    # magnitude (normalized) 
    mask[..., 2] = cv2.normalize(magnitude, None, 0, 255, cv2.NORM_MINMAX) 
      
    # Converts HSV to RGB (BGR) color representation 
    rgb = cv2.cvtColor(mask, cv2.COLOR_HSV2BGR) 

    # Normalize x and y components
    x_flow = cv2.normalize(flow[...,0], None, 0, 255, cv2.NORM_MINMAX)     
    y_flow = cv2.normalize(flow[...,1], None, 0, 255, cv2.NORM_MINMAX)
    x_flow = x_flow.astype('uint8')
    y_flow = y_flow.astype('uint8')

    # Change - Make next frame previous frame
    prvs = next_.copy()
    
    # Save each image
    cv2.imwrite(path + "x/x%s.jpg" % i, x_flow)
    cv2.imwrite(path + "y/y%s.jpg" % i, y_flow)
    cv2.imwrite(path + "rgb/rgb%s.jpg" % i, rgb)
    i = i+1

    # Increment counter to go to next frame
    counter += 1
    
print("End processing!")

[CHI-YU-SUNG]

useful, thanks

[guuzaa]

Hi @yangwangx ,
What's the version of your opencv library ?