how can i write video using matplotlib.plot images to Opencv Library

user_19210991 0
Viewed 88
    x = []
    y = []

    fourcc = cv2.VideoWriter_fourcc('M','J','P','G')
    out = cv2.VideoWriter("/Users/hyeonseuk/Desktop/ax.avi", fourcc, 30, (1000,200))

    def figure_to_array(fig):
        fig.canvas.draw()
        return np.array(fig.canvas.renderer._renderer)

    while len(x_data)>60+cnt:
        x = [i for i in range(cnt,cnt+60)]
        y = x_data[cnt:cnt+60,0]
        cnt+=1
        f = plt.figure(figsize=(5,1))
        plt.ylim(0,1)
        plt.plot(x,y)
        plt.close()
        f_arr = figure_to_array(f)
        f_arr = cv2.resize(f_arr,(1000,200))
        out.write(f_arr)
out.release()
cv2.destroyAllWindows()

this code is converting plot_image(f) to array(f_arr) and using this array writing video with opencv. convert plot to array is done. but using convert array with opencv to make video was written but not readable. how i can fix this problem?

Thank you.

matplotlib video opencv
1 Answers

f_arr is returned in RGBA pixel format, we have to convert it to BGR pixel format before writing to the video file.

RGBA has 4 color channels, and cv2.VideoWriter supports only 3 color channels, so we are getting corrupted video file.

Replace out.write(f_arr) with:

bgr = cv2.cvtColor(f_arr, cv2.COLOR_RGBA2BGR)
out.write(bgr)

Complete code sample (using some arbitrary values of x_data):

import cv2
import numpy as np
from matplotlib import pyplot as plt

x = []
y = []

x_data = np.vstack((np.arange(0, 100), np.arange(0, 100))).T  # arbitrary values (for testing).
cnt = 10

fourcc = cv2.VideoWriter_fourcc('M','J','P','G')
out = cv2.VideoWriter("ax.avi", fourcc, 30, (1000,200))

def figure_to_array(fig):
    fig.canvas.draw()
    return np.array(fig.canvas.renderer._renderer)

while len(x_data)>60+cnt:
    x = [i for i in range(cnt,cnt+60)]
    y = x_data[cnt:cnt+60,0]
    cnt+=1
    f = plt.figure(figsize=(5,1))
    #plt.ylim(0,1)
    plt.plot(x,y)
    #plt.show(block=True) # Show plot for testing
    plt.close()    
    f_arr = figure_to_array(f)
    f_arr = cv2.resize(f_arr,(1000,200))
    #cv2.imshow('f_arr', f_arr) # Show f_arr for testing
    #cv2.waitKey(10)
    bgr = cv2.cvtColor(f_arr, cv2.COLOR_RGBA2BGR)
    out.write(bgr)

out.release()
#cv2.destroyAllWindows()

Sample output video frame:
enter image description here

user_49267570
Related