How to detect and draw FAST feature points in OpenCV Python?

FAST (Features from Accelerated Segment Test) is a high-speed corner detection algorithm designed for real-time applications. OpenCV provides a simple interface to detect corner features using the FAST algorithm through cv2.FastFeatureDetector_create().

How FAST Algorithm Works

FAST detects corners by examining a circle of 16 pixels around each candidate point. If a continuous arc of pixels (usually 12 or more) are all brighter or darker than the center pixel by a threshold value, it's classified as a corner feature.

Steps to Detect FAST Features

To detect and draw feature points using the FAST detector, follow these steps ?

• Import the required libraries OpenCV. Make sure you have already installed it.

• Read the input image using cv2.imread() and convert it to grayscale using cv2.cvtColor().

• Create a FAST detector object with cv2.FastFeatureDetector_create().

• Detect keypoints using fast.detect(gray, None) method.

• Draw the detected keypoints using cv2.drawKeypoints() function.

• Display the result image with keypoints.

Example 1: FAST with Non-Max Suppression Enabled

In this example, we detect corner features with the default non-max suppression enabled ?

import cv2
import numpy as np

# Create a sample image with corners
img = np.zeros((400, 400, 3), dtype=np.uint8)
img[50:150, 50:150] = [255, 255, 255]  # White square
img[200:300, 200:300] = [128, 128, 128]  # Gray square
img[250:350, 50:150] = [200, 200, 200]  # Light gray square

# Convert to grayscale
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)

# Create FAST detector with default values
fast = cv2.FastFeatureDetector_create()

# Detect keypoints
keypoints = fast.detect(gray, None)

# Draw keypoints on the image
img_with_keypoints = cv2.drawKeypoints(img, keypoints, None, color=(0, 255, 0))

# Print detector parameters
print("Threshold:", fast.getThreshold())
print("Non-max Suppression:", fast.getNonmaxSuppression())
print("Neighborhood type:", fast.getType())
print("Total keypoints with non-max suppression:", len(keypoints))

Threshold: 10
Non-max Suppression: True
Neighborhood type: 2
Total keypoints with non-max suppression: 16

Example 2: FAST with Non-Max Suppression Disabled

Here we disable non-max suppression to detect more keypoints ?

import cv2
import numpy as np

# Create the same sample image
img = np.zeros((400, 400, 3), dtype=np.uint8)
img[50:150, 50:150] = [255, 255, 255]  # White square
img[200:300, 200:300] = [128, 128, 128]  # Gray square
img[250:350, 50:150] = [200, 200, 200]  # Light gray square

# Convert to grayscale
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)

# Create FAST detector and disable non-max suppression
fast = cv2.FastFeatureDetector_create()
fast.setNonmaxSuppression(False)

# Detect keypoints
keypoints = fast.detect(gray, None)

# Draw keypoints
img_with_keypoints = cv2.drawKeypoints(img, keypoints, None, color=(0, 0, 255))

# Print parameters
print("Threshold:", fast.getThreshold())
print("Non-max Suppression:", fast.getNonmaxSuppression())
print("Neighborhood type:", fast.getType())
print("Total keypoints without non-max suppression:", len(keypoints))

Threshold: 10
Non-max Suppression: False
Neighborhood type: 2
Total keypoints without non-max suppression: 64

Understanding Non-Max Suppression

Non-max suppression reduces the number of detected keypoints by removing weaker responses near stronger ones. When disabled, FAST detects significantly more keypoints, which may include redundant or less reliable corner points.

Key Parameters

Conclusion

FAST is an efficient corner detection algorithm ideal for real-time applications. Use non-max suppression for cleaner results, or disable it when you need to detect all possible corner candidates.