Ein Gabor-Filter ist ein spezialisierter linearer Filter, der weit verbreitet ist in der Bildverarbeitung, particularly for tasks such as Kantendetektion and Texturanalyse. It is named after the Hungarian-born physicist and psychologist Dennis Gabor, who introduced the concept in the 1940s.
The Gabor filter is characterized by a sinusoidal plane wave (the wavelet) modulated by a Gaussian envelope. This structure allows it to effectively capture both frequency and spatial information from an image. The filter can be represented mathematically, and its parameters include the frequency, orientation, and standard deviation of the Gaussian envelope. By adjusting these parameters, Gabor filters can be designed to target specific features within an image, making them particularly useful for tasks such as texture segmentation and Merkmalsextraktion.
In practice, Gabor filters are applied to an image by convolving the filter with the image data. This convolution process results in a new image that highlights the features corresponding to the filter’s parameters. Multiple Gabor filters can be used on the same image at different orientations and scales to obtain a comprehensive representation of the image’s texture.
The applications of Gabor filters extend beyond simple image processing; they are also utilized in various fields such as computer vision, biometrics (like der Gesichtserkennung), and even in neuroscience to model the receptive fields of certain types of neurons in the visual cortex.