In this recipe, you will learn how to classify textures using Gabor filter banks with scikit-image's filter module's functions. Frequency and orientation are two key parameters of the Gabor filter, which detects the presence of a given frequency content in an image in a given direction around the ROI neighborhood. The Gabor kernel has both a real and an imaginary part, where the real part is used to filter images. The features to be used for (texture) classification are the mean and variance (often based on LSE) of a filtered image. The Gabor filter's impulse response is a product of a sinusoidal function and a Gaussian function, as shown in the following image:
![](https://static.packt-cdn.com/products/9781789537147/graphics/assets/d6d2f198-7785-4cb0-a4d1-e8fa4217bd0d.png)