Image Processing Reference
FIGURE 2 Original NL (left); NL with detected edges (right).
FIGURE 3 NL with edges (right); NL with detected lines (right).
Corner detection is done primarily for text spoting because text segments tend to contain
many distinct corners. Thus, image segments with higher concentrations of corners are likely
stages of the dilate-erode method with different 5 × 5 kernels are applied. Two stages of dilate-
erode with different kernels are applied. The first stage uses a 5 × 5 cross dilate kernel for hori-
zontal and vertical expansions. It then uses a 5 × 5 diamond erode kernel for diagonal shrink-
ing. The resulting image is compared with the original and those pixels which are in the corner
of an aligned rectangle are found.
FIGURE 4 Corner detection steps.
The second stage uses a 5 × 5 X-shape dilate kernel to expand in the two diagonal directions.
A 5 × 5 square kernel is used next to erode the image and to shrink it horizontally and ver-
tically. The resulting image is compared with the original and those pixels which are in a 45°
corner are identified. The resulting corners from both steps are combined into a final set of de-
mond kernels are used to detect aligned corners. The botom sequence of images corresponds
to stage two when the X-shape and square kernels are used to detect 45° corners. Step one
shows the original input of each stage, step two is the image after dilation, step three is the
image after erosion, and step four is the difference between the original and eroded versions.
The resulting corners are outlined in red (dark gray in print versions) in each step to provide
a basis of how the dilate-erode operations modify the input.