23.11.2012 Views

Master Thesis - Hochschule Bonn-Rhein-Sieg

Master Thesis - Hochschule Bonn-Rhein-Sieg

Master Thesis - Hochschule Bonn-Rhein-Sieg

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

5. Algorithms <strong>Master</strong> <strong>Thesis</strong> Björn Ostermann page 61 of 126<br />

used _ range � max_brightness<br />

� min_brightness<br />

complete _ range<br />

scaling _ factor �<br />

used _ range<br />

new_<br />

brightness � scaling _ factor �<br />

Equation 3: Scaling algorithm<br />

a) b)<br />

Number of Pixels<br />

Offset<br />

Used Range<br />

Complete Range<br />

Illumination<br />

c) d)<br />

Number of Pixels<br />

Offset<br />

Used Range<br />

Complete Range<br />

Illumination<br />

�brightness � min_brightness�<br />

Figure 38: Scaling algorithm a) to b) with a small used range<br />

and c) to d) with a large used range<br />

Number of Pixels<br />

Number of Pixels<br />

Illumination<br />

Illumination<br />

This pixel distribution of more than one information cluster is not a special, but a normal case,<br />

especially in industrial environments where lowly reflecting materials like cloth meet highly reflective<br />

materials like metals. In the case of the used workspace the mounting of the camera, visible in the<br />

camera’s image, has a high reflectivity, whereas the floor has a very low one. Thus clusters at both<br />

ends of the intensity range are always present.<br />

This problem can be avoided by using the histogram equalization, which is described in Figure 39,<br />

starting with the real distribution in Figure 39a and ending with the resulting distribution in Figure<br />

39e. In the real distribution, there are two regions visible in which the majority of the pixels’ values<br />

are located.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!