Image processing is a vast field that encompasses a wide range of techniques for modifying digital images. One of the most important processing techniques is filtering. Filtering refers to the process of applying one or more mathematical operations to each pixel in an image in order to achieve a desired effect.
One of the most common types of filters used in image processing is the high pass filter. A high pass filter is a mathematical operation that enhances the highfrequency components of an image while suppressing the lowfrequency components.
In this article, we will discuss the concept of high pass filters in image processing, their importance, and how they can be implemented using code examples.
What is a High Pass Filter?
A high pass filter is a type of filter that allows highfrequency components of an image to pass while attenuating or blocking lowfrequency components. In image processing, the highfrequency components refer to edges and details in the image, while lowfrequency components refer to the gradual changes in the intensity values of an image.
The high pass filter enhances the edges and details in an image by amplifying the highfrequency components while suppressing the lowfrequency components. This results in an image that appears sharper and clearer.
Why High Pass Filters are Important in Image Processing?
High pass filters are essential tools in image processing as they help to sharpen and enhance the edges and details in an image. They are particularly useful in medical imaging, where it is essential to enhance the edges of tissues and organs for accurate diagnosis.
High pass filters are also used in video processing to sharpen and enhance the details in video frames. They are also commonly used in photography to enhance the sharpness and clarity of digital images.
How to Implement High Pass Filters in Image Processing?
High pass filters can be implemented in several ways in image processing. Some of the commonly used methods for implementing high pass filters are:
 Using Convolution
The most commonly used method for implementing high pass filters is by using convolution. Convolution is a mathematical operation that involves multiplying a matrix of values, called a kernel or a filter, with each pixel in an image. The result of this multiplication is then summed up to obtain the output value for that pixel.
The high pass filter is a type of convolution kernel that amplifies highfrequency components while suppressing lowfrequency components. The high pass filter kernel is defined as:
1/9 1/9 1/9
1/9 8/9 1/9
1/9 1/9 1/9
The above kernel is a 3×3 filter that amplifies the highfrequency components by subtracting the average intensity of the neighboring pixels. It then normalizes the output by dividing the result by 9 to ensure that the output values are within the range of 0 to 255.
The following code demonstrates how to implement a high pass filter using convolution in Python:
import cv2
import numpy as np
img = cv2.imread('image.jpg')
Define a high pass filter kernel
kernel = np.array([[1,1,1],
[1,9,1],
[1,1,1]])
Apply the high pass filter using convolution
output = cv2.filter2D(img, 1, kernel)
Display the original and filtered image
cv2.imshow('Original Image', img)
cv2.imshow('High Pass Filtered Image', output)
cv2.waitKey(0)
cv2.destroyAllWindows()
 Using FFT
Another method for implementing high pass filters is by using Fourier transforms. The Fourier transform is a mathematical operation that converts a signal from the time domain to the frequency domain.
In image processing, the Fourier transform is used to analyze the frequency components of an image. By using Fourier transforms, it is possible to separate the low and high frequency components of an image.
The high pass filter can be implemented using Fourier transforms by first converting the image to the frequency domain and then removing the lowfrequency components to amplify the highfrequency components.
The following code demonstrates how to implement a high pass filter using Fourier transforms in Python:
import cv2
import numpy as np
from matplotlib import pyplot as plt
img = cv2.imread('image.jpg',0)
Perform a Fourier transform
f = np.fft.fft2(img)
Shift the zerofrequency component to the center of the spectrum
fshift = np.fft.fftshift(f)
Set the radius of the central circle to 20% of the image height or width, whichever is smaller
rows, cols = img.shape
crow,ccol = int(rows/2) , int(cols/2)
r = min(rows,cols)/5
fshift[crowr:crow+r, ccolr:ccol+r] = 0
Shift the zerofrequency component back to the topleft corner of the spectrum
f_ishift = np.fft.ifftshift(fshift)
Perform an inverse Fourier transform
output = np.fft.ifft2(f_ishift)
output = np.abs(output)
Display the original and filtered image
plt.subplot(121),plt.imshow(img, cmap = 'gray')
plt.title('Original Image'), plt.xticks([]), plt.yticks([])
plt.subplot(122),plt.imshow(output, cmap = 'gray')
plt.title('High Pass Filtered Image'), plt.xticks([]), plt.yticks([])
plt.show()
Conclusion
In conclusion, high pass filters are essential tools in image processing that can help enhance the edges and details in an image by amplifying the highfrequency components while suppressing the lowfrequency components.
High pass filters can be implemented using convolution or Fourier transforms, and the choice of implementation depends on the specific needs of the application. With the examples provided above, you should now be able to implement high pass filters in your image processing projects.
let's dive deeper into some of the topics we covered earlier.
High Pass Filter Applications in Image Processing
High pass filters are used in a wide range of applications within image processing. These filters are commonly used to sharpen an image or to remove blurring caused by lowpass filters. High pass filters are also used in edge detection applications, where it is necessary to identify the edges of objects within an image, such as in medical imaging or in the analysis of satellite images.
Another area where high pass filters can be effectively applied is in the enhancement of low contrast images. Low contrast images are those in which the difference between the darkest and lightest regions is relatively small. Applying a high pass filter to these images can help to reveal previouslyhidden details and enhance the contrast between different regions of the image.
There are also specialized applications, such as in the field of forensic analysis, where high pass filters can be used to enhance security footage or images captured from surveillance systems. High pass filters can help reveal previouslyhidden details in these images, such as license plate numbers or facial features.
Code Example Using OpenCV in Python
In the implementation section above, we showed an example of how to implement a high pass filter using convolution in Python. In this section, we will provide a code example of how to implement a high pass filter using OpenCV, a popular opensource library for computer vision applications.
The following is a code example that demonstrates how to apply a high pass filter to an image using OpenCV in Python:
import cv2
import numpy as np
# Load an image
image = cv2.imread('example_image.jpg', 0)
# Define a high pass filter kernel
kernel = np.array([[1,1,1],
[1,9,1],
[1,1,1]])
# Apply the high pass filter using OpenCV's filter2D function
filtered_image = cv2.filter2D(image, 1, kernel)
# Display the original and filtered image
cv2.imshow('Original Image', image)
cv2.imshow('High Pass Filtered Image', filtered_image)
cv2.waitKey(0)
cv2.destroyAllWindows()
In the above code, we first load an image using OpenCV's imread function. We then define a high pass filter kernel of size 3×3, and apply the high pass filter to the loaded image using OpenCV's filter2D function. Finally, we display the original and filtered images using OpenCV's imshow function.
Conclusion
High pass filters are an essential tool in image processing and can be used in a wide range of applications to enhance the edges and details within an image. There are various methods for implementing high pass filters, including using convolution or Fourier transforms, and the choice of implementation depends on the specific requirements of the application.
In this article, we discussed the concept of high pass filters, why they are important in image processing, and provided code examples for how to implement high pass filters using both convolution and OpenCV in Python.
Popular questions

What is a high pass filter and how does it work in image processing?
A high pass filter is a type of filter that amplifies highfrequency components in an image while suppressing lowfrequency components. Highfrequency components typically refer to edges and details in an image while lowfrequency components refer to gradual changes in intensity values. The high pass filter works by removing the lowfrequency components through mathematical operations, resulting in a sharper, clearer image. 
What are some common applications of high pass filters in image processing?
High pass filters are commonly used for sharpening images, edge detection, enhancing lowcontrast images, and forensic analysis. They can also be used in video processing and photography to enhance the details and clarity of images. 
How can a high pass filter be implemented using convolution in Python?
A high pass filter can be implemented using convolution in Python by defining a kernel or filter that amplifies highfrequency components and suppresses lowfrequency components. This filter can then be applied using the filter2D function in OpenCV. 
Can high pass filters be applied in other programming languages besides Python?
Yes, high pass filters can be implemented in other programming languages besides Python. The mathematical operations used in high pass filters are standard, so the same operations can be implemented using other programming languages such as MATLAB, C++, or Java. 
Is there a difference between high pass filters implemented using convolution and those implemented using Fourier transforms?
Yes, there is a difference between high pass filters implemented using convolution and those implemented using Fourier transforms. Convolution is a process of multiplying pixel values with a kernel matrix, while Fourier transforms break down an image into frequency values. Convolutionbased high pass filters work directly with the pixels in an image, while Fourier transformbased high pass filters work in the frequency domain. The choice of implementation depends on the specific requirements of the application.
Tag
Sharpness.