Читать книгу Rethinking Prototyping - Группа авторов - Страница 120

3.4 Colour Space

Оглавление

Fig. 6 RGB colour cube with image point cloud and bounding box

After the values are Fourier transformed, filtered and then inverse transformed again, they are not in the same absolute range anymore but only keep their relative proportion. In order to again achieve the same chromaticity of the input image, the values have to be mapped to the original colour space. For that reason, the minimum and maximum value of each channel (red, green and blue) are stored. These limits describe a rhomboid within the entire colour cube. This rhomboid is the bounding box of the point cloud, where each pixel represents a point defined by the channels as its coordinates.

In a grey scale image, each pixel’s colour value has identical red, green and blue content. They all would sit on the diagonal line going from black (0/0/0) to white (255/255/255). Because of the linear dependence, an alternative description could be therefore


Rethinking Prototyping

Подняться наверх