Animation #1 - Simple animation using numpy

Yesterday, I wrote a blog post on rotating an image with OpenCV. I will explain how to create a simple animation using the rotate and pan functions of an image. I hope this learning will give you a deeper opportunity to learn more about OpenCV and numpy.

Basic concept

Add an overlay image to the right of the canvas image. As time goes, rotate the overlay image and move it to the left.

You need a bit of math knowledge to create natural animations.

Angular Velocity

Angular velocity refers to how fast an object rotates or revolves relative to another point, i.e. how fast the angular position or orientation of an object changes with time. I'm going to make an 30 FPS(frame per second)  animation. Thus, the time per frame is 1/30 seconds (0.0333). If it takes 2 seconds to rotate the image 360 ​​degrees, the angular velocity is π(radian)/sec. Thus, the angle traveled for 1/30 second is π(radian) / 30, and the distance that the circle rolled is R x  π(radian) / 30.

<image from https://www.shelovesmath.com/trigonometry/linear-and-angular-speeds-area-of-sectors-and-length-of-arcs/>

Let's summarize.

• FPS = 30
• Angular Velocity(radian) = π(radian) / sec
• Angular Velocity(degree) = 180(degree) / sec
• step_angle(radian) = π(radian) / FPS
• step_angle(degree) = 180(degree) / FPS
• step_movement(pixel) = step_angle(radian) * R = step_angle(radian) * (h/2)

Make animation

import argparse
import cv2
import numpy as np
import time

FPS = 30
angular_velocity = np.degrees(np.pi)    # I'll make 1 rotation per 2 seconds
step_angle = angular_velocity / FPS
step_radian = np.radians(step_angle)
max_count = 3
def process_masking(base, mask, pos):
    h, w, c = mask.shape
    hb, wb, _ = base.shape
    x = pos[0]
    y = pos[1]

    #check mask position
    if(x > wb or y > hb):
        print(' invalid overlay position(%d,%d)'%(x, y))
        return None
    
    #remove alpha channel    
    if c == 4:
        mask = cv2.cvtColor(mask, cv2.COLOR_BGRA2BGR) 
    
    #adjust mask
    if(x + w > wb):
        mask = mask[:, 0:wb - x]
        print(' mask X size adjust[W:%d] -> [W:%d]'%(w, wb - x))
    if(y + h > hb):
        mask = mask[0:hb - y, :]
        print(' mask Y size adjust[H:%d] -> [H:%d]'%(h, hb - y))

    h, w, c = mask.shape
    
    img = base.copy()
    bg = img[y:y+h, x:x+w]      #overlay area
    try:
        for i in range(0, h):
            for j in range(0, w):
                B = mask[i][j][0]
                G = mask[i][j][1]
                R = mask[i][j][2]
                if (int(B) + int(G) + int(R)):
                    bg[i][j][0] = B
                    bg[i][j][1] = G
                    bg[i][j][2] = R
        img[y:y+h, x:x+w] = bg
    except IndexError:  #index (i, j) is out of the screen resolution.  (화면 범위를 벗어남.)
        print(' index Error')
        return None
    return img

def delay_fps(s):
    while (time.time() - s < (1.0 / FPS) ):
        time.sleep(0.001)

parser = argparse.ArgumentParser(description="OpenCV Example")
parser.add_argument("--file", type=str, required=True, help="filename of the input image to process")
args = parser.parse_args()


mask = cv2.imread(args.file, cv2.IMREAD_COLOR)
height, width, channels = mask.shape
print("image   H:%d W:%d, Channel:%d"%(height, width, channels))
print('Angular Velocity:%f step_angle:%f'%(angular_velocity, step_angle))
canvas = np.zeros((height * 2, width * 5, 3), np.uint8)
c_height, c_width, c_channels = canvas.shape

fourcc = cv2.VideoWriter_fourcc('m', 'p', '4', 'v')
out_video = cv2.VideoWriter('./animation.mp4', fourcc, 30, (c_width, c_height))


angle = step_angle
x_pos = (c_width - width)*1.0
count = 0

while count < max_count:
    matrix = cv2.getRotationMatrix2D((width/2, height/2), angle, 1)
    rotate = cv2.warpAffine(mask, matrix, (width, height))
    s = time.time()
    angle += step_angle
    if(angle > 360):
        angle = angle % 360

    print('x_pos:%f'%(x_pos))
    img = process_masking(canvas, rotate, (int(x_pos),0))
    x_pos -= step_radian * height / 2.0 
    print('x_pos:%f'%(x_pos))
    out_video.write(img)
    cv2.imshow('rotate', img)
    if(x_pos < count * width):
        x_pos = (c_width - width)*1.0
        count += 1
        canvas = img.copy()
    k = cv2.waitKey(1)
    delay_fps(s)
    

out_video.release()
cv2.waitKey(0)
cv2.destroyAllWindows()

Run the code.

python animation_rotate.py --file=sbear.png
image   H:64 W:64, Channel:3
Angular Velocity:180.000000 step:6.000000

You can get a animation file like this.

You can download the source codes here(https://github.com/raspberry-pi-maker/OpenCV)