.
FoveaPro Image Processing

Everything from edge detection to halftone removal, texture extraction to dynamic range enhancement...

Fovea Pro has an extensive set of image processing capabilities, for both 8 and 16 bit per channel greyscale and RGB color images. This includes neighborhood processing (in the spatial domain) and frequency processing (in the Fourier domain). Spatial domain operations include convolutions, rank-based operations, as well as edge, orientation and texture enhancement tools. A few illustrative examples are shown here.

Rank operator example

Fovea Pro's versatile rank operator removes noise, fits background, and extracts texture and edges.


Convolution example

Fovea Pro can apply any convolution kernel (containing real or integer values) up to 95x95 in size.


Difference of gaussians example

More versatile than an Unsharp Mask! The Difference of Gaussians filter allows both edge enhancement and noise reduction. Fovea Pro also includes adaptive neighborhood equalization and rank-based operators for image sharpening.


Texture example

Fovea Pro can extract texture and orientation (the image shows a flycut metal surface processed to distinguish the lay or machining direction).


Edge finding example

Edge enhancement operations include both the common ones (Sobel, Kirsch, etc.) and the more advanced ones (Frei & Chen, Canny, etc.) that are rarely found outside of dedicated imaging systems.


Fourier example

There are many ways to process an image. Processing on the pixels alone is insufficient in some cases. Fourier space processing is particularly good for removing periodic noise. Fovea Pro includes both manual and automatic ways to generate Fourier-space filters.


Removing halftoning with Fourier processing

A half-tone image (newspaper picture) and the result of properly removing the halftone pattern with a Fourier-space mask. This is a much more powerful method than blurring the image, which removes fine detail.


Using Fourier deconvolution to sharpen images

The use of a 16 bit image depth is especially important for image sharpening using deconvolution in Fourier space. Fovea Pro has a complete set of FFT tools for processing and analysis.


Adjustable deconvolution setting.

Here you can see the PSF (Point Spread Function) and the image being deconvolved. The Weiner-K Noise Level is used to adjust the sharpness of the result.

Next: Principal Components Analysis
Prev: Image Adjustment
Up: Fovea Pro

 

.