Feature extraction approach in Computer Vision - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to extract features from an image using a simple method.

Computer Vision

import cv2
image = cv2.imread('image.jpg', 0)
features = cv2.Canny(image, [1], 150)
print(features.shape)

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

A50

B100

C200

D300

Attempts:

3 left

💡 Hint

Common Mistakes

Using a threshold too high or too low can miss edges or detect too many.

The Canny edge detector requires two threshold values. Here, 100 is a common lower threshold to detect edges.

2fill in blank

medium

Complete the code to compute Histogram of Oriented Gradients (HOG) features from an image.

Computer Vision

from skimage.feature import hog
from skimage import data, exposure
image = data.astronaut()
features, hog_image = hog(image, pixels_per_cell=(8, 8), cells_per_block=(2, 2), visualize=True, [1]=True)
print(features.shape)

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

Amultichannel

Bnormalize

Cblock_norm

Dtransform_sqrt

Attempts:

3 left

💡 Hint

Common Mistakes

Forgetting to set multichannel=True causes errors or wrong features.

The 'multichannel=True' parameter tells the function to treat the input as a color image.

3fill in blank

hard

Fix the error in the code to extract SIFT features from an image.

Computer Vision

import cv2
image = cv2.imread('image.jpg', 0)
sift = cv2.SIFT_create()
keypoints, descriptors = sift.[1](image)
print(len(keypoints))

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

AdetectAndCompute

Bdetect

Ccompute

DdetectAndDescribe

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'detect' or 'compute' alone causes errors or incomplete results.

The correct method to detect keypoints and compute descriptors in one step is 'detectAndCompute'.

4fill in blank

hard

Fill both blanks to create a dictionary of feature descriptors for each keypoint.

Computer Vision

features_dict = [1](kp.pt: desc for kp, desc in zip(keypoints, [2]))
print(len(features_dict))

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

Adict

Bdescriptors

Clist

Dkeypoints

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'list' instead of 'dict' or wrong variable names.

We use 'dict' to create a dictionary, and 'descriptors' holds the feature vectors for each keypoint.

5fill in blank

hard

Fill all three blanks to filter features with descriptor length greater than 100.

Computer Vision

filtered_features = {kp: desc for kp, desc in features_dict.items() if len(desc) [1] [2]
print(len(filtered_features))

threshold = [3]

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

A>

B100

C<

D50

Attempts:

3 left

💡 Hint

Common Mistakes

Using '<' instead of '>' or wrong threshold values.

The code filters descriptors with length greater than 100, so the operator is '>' and threshold is 100.