How to detect a triangle in an image using OpenCV Python?

When it comes to computer vision, OpenCV is one of the most utilized libraries available for Python developers. It encompasses a vast array of computer vision, image processing, and machine learning concepts, including edge detection, feature recognition, object detection, and much more. In this article, we explore how to detect a triangle in an image with OpenCV.

Pre-requisites

Before we dive into the code, let us first understand some of the key concepts. We will be using the following libraries:

• NumPy
• OpenCV

Steps to Detect a Triangle in an Image

The steps involved in detecting a triangle in an image using OpenCV are:

1. Load the image
2. Convert the image to grayscale
3. Apply Gaussian blur
4. Perform edge detection using the Canny edge detector
5. Find contours in the image
6. Iterate over the contours and check if each contour is a triangle

Let us now dive into the code for each of these steps.

Load the image

We begin by loading the image.

import cv2
import numpy as np

image = cv2.imread("image.png")

Note that we are making use of the cv2.imread function to load the image. The image should be present in the same directory as the Python script.

Convert the image to grayscale

Converting the image to grayscale simplifies the image processing operations.

gray_image = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)

Here, we are using the cv2.cvtColor function to convert the image from BGR to grayscale.

Apply Gaussian blur

Gaussian blur is a common technique used to reduce noise in images. It essentially smoothes the image by averaging the pixel values in a kernel of a certain size around each pixel.

blurred_image = cv2.GaussianBlur(gray_image, (5, 5), 0)

Here, we are using the cv2.GaussianBlur function to apply Gaussian blur to the grayscale image. The kernel size (5, 5) and standard deviation (0) are parameters that need to be adjusted based on the image being processed.

Perform edge detection using the Canny edge detector

The Canny edge detector is a widely used technique to detect edges in images. It works by identifying the intensity gradients in the image and highlighting regions with high gradients.

canny_edges = cv2.Canny(blurred_image, 50, 150)

Here, we are using the cv2.Canny function to apply edge detection to the blurred image. The parameters 50 and 150 represent the minimum and maximum intensity gradients.

Find contours in the image

Contours are simply the boundaries or curves that join the continuous points of an image with the same intensity or color. In this step, we identify all the contours present in the image.

contours, hierarchy = cv2.findContours(canny_edges, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)

We make use of the cv2.findContours function to get the contours present in the image. The parameters cv2.RETR_EXTERNAL and cv2.CHAIN_APPROX_SIMPLE are used to retrieve only the external contours and approximate the contours, respectively.

Iterate over the contours and check if each contour is a triangle

Finally, we loop over all the contours and check if each contour is a triangle. A contour is considered to be a triangle if it has three edges.

for cnt in contours:
    approx = cv2.approxPolyDP(cnt, 0.01 * cv2.arcLength(cnt, True), True)

    if len(approx) == 3:
        cv2.drawContours(image, [cnt], 0, (0,255,0), 3)
        print("Triangle Detected")

In the above code, we are iterating over all the contours present in the image. Next, we use the cv2.approxPolyDP function to approximate the contours as straight lines with an accuracy of 0.01 times the contour perimeter. If the number of edges is 3, we can conclude that the contour is a triangle. Finally, we draw the triangle contours on the original image using the cv2.drawContours function.

Conclusion

In this article, we have explored how to detect a triangle in an image using OpenCV in Python. We have covered the steps involved in detecting a triangle, including loading the image, converting it to grayscale, applying Gaussian blur, performing edge detection using the Canny edge detector, finding contours in the image, and checking if each contour is a triangle. Armed with this knowledge, you can apply similar techniques to detect other shapes or objects in images.