(GO) 3x3 convolution kernels with online demo3x3 convolution kernels with online demo
- Which is the most used 3x3 convolution kernels/matrices?
- Which kernel is used for averaging, applying blur or smooth effect, does sharpening or for the emboss effect?
- Which kernels can is used to detect edges, calculate the gradient or the smoothed gradient?
- Can I Try somewhere the dimensional convolution in an interactive application?
The frequently used 3x3 convolution kernels is listed below with some short description. At the end of this post there is a interactive demo, where you can try and play with different 3x3 kernels. For the mathematical background of the usage of these kernels, please read the previous post on both dimensional convolutio N.
Average (blur, smooth) 3x3 convolution kernel
This kernel was used for noise reduction and blurring the image. Must is normalized, otherwise the result may not fit the (0, 255) range.
Sharpen 3x3 convolution kernel
This kernel was used to enhance the small differences and edges in the image.
Edge Detection 3x3 convolution kernels
These kernels is sensitive to the edges. Kernel E is for detecting in both directions, while EH and EV be sensitive for the horizontal and vertical edges respectively.
Gradient Detection 3x3 convolution kernels
Kernels GH and GV is to calculate the magnitude of the horizontal and the vertical gradient.
Sobel operator 3x3 convolution kernels
Sobel operators is similar to the gradient kernels approximating the smoothed gradient of the image in horizontal and ver Tical directions. It can be seen from the decomposing, which this operator is a combination of a gradient detector and a smoothing kernel.
Emboss 3x3 Convolution kernel
This kernel creates a embossing effect, can be rotated to modify the direction of this operator.
Demo Application
This program demonstrates the using 3x3 convolution kernels on classic image processing source images. Click on the label to load the application. It shall run in every modern browser, including ie9+.
Demo app See original
You can use the currently filtered image as source by clicking the use filtered image button. Some predefined kernels can is chosen, but the values are directly selectable too. Setting Filter Normalization divides the convolution result by the summary of the elements in the kernel. It's important to mention, which on the result image the absolute output values is shown.
3*3 Convolution Verification Example