Images are an integral part of our daily lives. They convey information, emotions, and art.
With the advent of technology, we can now manipulate images using software to suit our preferences. Image filters are a popular way to transform images, and in this article, we will explore different types of filters and learn how to apply them.
Importing Modules
Before we begin exploring filters, we need to import relevant modules to our code. These modules include OpenCV, matplotlib, and numpy.
OpenCV is a traditional library used for image processing. matplotlib is an excellent visualization tool, while numpy is used for numerical computations.
We also need to set the plotting style to font_sans-serif to give our graph a more professional look.
Loading the initial image
Now that we have imported the modules, the next step is loading the initial image. We can load the image using the imread function in OpenCV.
The imread function reads an image from the specified file and returns an array of image contents in color format. We can also use cvtColor to convert the image’s color format to black and white or grayscale.
Finally, we can use imshow to display the image.
Creating Filters
To apply filters, we first need to define the filter as a kernel. Kernels are matrices of odd sizes that we can use to perform mathematical operations on the image.
The 2D filters perform operations on adjacent pixels. After defining the kernel, we apply it using the filter2D function in OpenCV.
1. Emboss Filter
The emboss filter is used to give an image an engraved or 3D look. This filter is defined as a kernel that highlights edges in an image.
The matrix used is shown below:
-2 -1 0
-1 1 1
0 1 2This kernel multiplies the current pixel and its neighbors by the matrix values and adds the result to the output image. The result highlights the edges and gives the image a three-dimensional texture.
2. Sharpening Filter
The sharpening filter is used to enhance an image’s features. This filter is defined as a kernel that subtracts the current pixel value from its neighbor’s pixel value.
The matrix values are shown below:
0 -1 0
-1 5 -1
0 -1 0The kernel multiplies the current pixel and its neighbors by the matrix values and adds the result to the output image. The result is a sharper and more defined image.
3. Sepia Filter
The sepia filter applies a warm, brown hue to an image to give it an antique look. This filter is defined as a kernel that transforms the RGB image to sepia values of red, green, and blue.
The matrix values are shown below:
0.393 0.769 0.189
0.349 0.686 0.169
0.272 0.534 0.131The kernel multiplies the current pixel and its neighbors by the matrix values and adds the result to the output image. The result is a warm-toned image that gives it a classic look.
4. Blur Filter
The blur filter reduces the noise in an image by removing high-frequency content. This filter is defined as a kernel that smooths out edges in an image.
The matrix values are shown below:
1/9 1/9 1/9
1/9 1/9 1/9
1/9 1/9 1/9The kernel multiplies the current pixel and its neighbors by the matrix values and adds the result to the output image. The result is a softer image with fewer sharp edges.
Conclusion
In conclusion, image filters are a powerful tool for transforming images. These filters can give an image a unique look and feel depending on the filter applied.
We have learned about different filters, including the emboss filter that highlights edges, the sharpening filter that enhances image features, the sepia filter that gives an antique look, and the blur filter that smooths out edges in an image. By applying these filters, we can enhance the aesthetics and emotion conveyed by images.
In the previous sections, we discussed the different types of filters and how they can be applied to images to achieve various effects.
In this section, we will examine the final output of applying image filters to images. We will also summarize the key points that we have covered in this article.
Displaying all the filters to images
Now that we have explored various filters and learned how to build and apply them, let us display the output images after applying each filter. We will use the same initial image for each filter so that we can compare the differences effectively.
1. Emboss Filter Output Image
After applying the emboss filter to the initial image, the output image appears to have an engraved or 3D look. The edges become more pronounced and show more details.
The emboss filter enhances the texture in the image and makes it appear more lifelike.
2. Sharpening Filter Output Image
After applying the sharpening filter to the initial image, the output image becomes much more defined and sharper. The details of the image become more apparent.
The sharpening filter is particularly useful for images that need precise details or for images that are slightly blurred.
3. Sepia Filter Output Image
After applying the sepia filter to the initial image, the output image gains a warm, vintage feel. The sepia filter gives the image an antique look and makes it ideal for framing or displaying in a vintage photo album.
Applied to the right image, the sepia filter works wonders in adding dimension and character to the shot.
4. Blur Filter Output Image
After applying the blur filter to the initial image, the output image becomes softer, and the edges lose their sharpness. The blur filter is useful for reducing image noise and hide minor imperfections in the image.
The blur filter can make an image appear more tranquil and serene.
Conclusion
In conclusion, image filters are a powerful tool for manipulating images to achieve various effects. By applying different types of filters such as the emboss filter that highlights edges, the sharpening filter that enhances image details, the sepia filter that gives an antique look and the blur filter that reduces noise, we can drastically alter the appearance of an image.
We learned how to build and apply these filters to achieve the desired effect. It is important to remember that image filters do not work in all cases.
They work best on certain types of images, and the user should have a clear goal in mind before applying them. Finally, let’s bear in mind that filters are tools and that there are limitless ways to build and apply them.
Experimenting with different filters and techniques can result in fantastic images and creative original work. In this article, we delved into the world of image filters, learning about the different types of filters such as the emboss, sharpening, sepia, and blur filters.
We discussed how to build and apply these filters using OpenCV, matplotlib, and numpy, among other modules. By doing so, we were able to transform images, giving them a unique look that suits our preference.
It is important to remember that filters do not work in all cases, but experimenting with different filters and techniques can lead to fantastic images. In all, image filters are tools that can enhance the aesthetics and emotion conveyed by images.