Filter Image Concepts

Image Filters serve a variety of purposes, such as reducing noise, detecting edges along a specific direction, contouring patterns, and detail outlining or smoothing. Filters can smooth, sharpen, transform, and remove noise from an image so that you can extract the information you need. Filter algorithms are divided into two types: linear and nonlinear.

A linear filter, or convolution, is an algorithm that consists of recalculating the value of a pixel based on its own pixel value and the pixel values of its neighbors weighted by the coefficients of a convolution kernel. The sum of this calculation is divided by the sum of the elements in the kernel to obtain a new pixel value. Vision Builder AI features a set of standard convolution kernels for the most common sizes (3×3, 5×5, and 7×7). You can also create your own convolution kernels. With this capability, you can create filters with specific characteristics.

Nonlinear filters either extract the contours (edge detection) or remove the isolated pixels. Vision Builder AI has four different methods you can use for contour extraction (Differentiation, Prewitt, Roberts, or Sobel).

Filters alter pixel values with respect to variations in light intensity in their neighborhood. The neighborhood of a pixel is defined by the size of a matrix, or mask, centered on the pixel itself. These filters can be sensitive to the presence or absence of light-intensity variations.