Optical flow concept in Computer Vision - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to calculate the optical flow using Lucas-Kanade method.

Computer Vision

flow = cv2.calcOpticalFlow[1](prev_gray, next_gray, None, lk_params['winSize'], lk_params['maxLevel'], lk_params['criteria'])

Drag options to blanks, or click blank then click option'

APyrLK

BFarneback

CSparse

DDense

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'Farneback' instead of 'PyrLK' for sparse optical flow.

Confusing dense and sparse optical flow methods.

The Lucas-Kanade method is called using cv2.calcOpticalFlowPyrLK in OpenCV.

2fill in blank

medium

Complete the code to convert a video frame to grayscale for optical flow calculation.

Computer Vision

gray_frame = cv2.cvtColor(frame, [1])

Drag options to blanks, or click blank then click option'

Acv2.COLOR_RGB2GRAY

Bcv2.COLOR_BGR2RGB

Ccv2.COLOR_BGR2GRAY

Dcv2.COLOR_GRAY2BGR

Attempts:

3 left

💡 Hint

Common Mistakes

Using RGB conversion instead of grayscale.

Not converting the frame to grayscale before optical flow.

Optical flow algorithms require grayscale images, so we convert BGR frames to grayscale using cv2.COLOR_BGR2GRAY.

3fill in blank

hard

Fix the error in the code to calculate the magnitude and angle of optical flow vectors.

Computer Vision

magnitude, angle = cv2.cartToPolar(flow[..., 0], flow[..., [1]], angleInDegrees=True)

Drag options to blanks, or click blank then click option'

A1

B3

C2

D0

Attempts:

3 left

💡 Hint

Common Mistakes

Using index 2 or 3 which are out of bounds.

Using the same index 0 for both components.

The optical flow has two channels: horizontal (index 0) and vertical (index 1). We use index 1 for the vertical component.

4fill in blank

hard

Fill both blanks to create a mask image that visualizes optical flow direction and magnitude.

Computer Vision

mask = np.zeros_like(frame)
mask[..., [1]] = angle / 2
mask[..., [2]] = 255

Drag options to blanks, or click blank then click option'

A0

B1

C2

D3

Attempts:

3 left

💡 Hint

Common Mistakes

Confusing channel indices for HSV color space.

Setting saturation to zero instead of 255.

Hue channel is at index 0 and saturation channel is at index 1 in HSV images.

5fill in blank

hard

Fill all three blanks to convert the mask to BGR color and display it with OpenCV.

Computer Vision

rgb_flow = cv2.cvtColor(mask, [1])
cv2.imshow('Optical Flow', rgb_flow)
cv2.waitKey([2])
cv2.destroyAllWindows()

# Wait time of [3] milliseconds allows frame update

Drag options to blanks, or click blank then click option'

Acv2.COLOR_BGR2HSV

Bcv2.COLOR_HSV2BGR

Ccv2.COLOR_RGB2BGR

Dcv2.COLOR_GRAY2BGR

E0

F1

G30

H100

Attempts:

3 left

💡 Hint

Common Mistakes

Using wrong color conversion codes.

Using waitKey(1) or waitKey(100) which may be too fast or slow.

We convert HSV mask to BGR for display using cv2.COLOR_HSV2BGR. A waitKey of 0 waits indefinitely, but 30 ms allows smooth video playback.