IMAGEDATA - a structure to describe image contents

The structure describe pixel addresses in an image, so that the pixels are accessible to other algorithms. The structure is frequently used in graphical operations and image processing applications (like the FOuCUS and the FourierAnalyzer).

The C/C++ definition is the following:

typedef struct { unsigned char *Address; // the address of the top-left pixel int Type, // pixel type Width, // image width Height, // image height hPitch, // the difference between the vertical pixels (in bytes) vPitch; // -"- horizontal pixels (that is, the lines) } IMAGEDATA;

The meaning of the structure is the following:

With this, a pixel address can be calculated:

Address = id.Address + id.hPitch*x + id.vPitch*y

The structure can easily be filled to describe a Device Independent Bitmap (DIB) on Windows, but other libraries are not problematic, either.

Properties, examples:

• In case of 8 bit (grayscale) images, the hPitch is tipically 1.
• The structure can be used o process 24 bit (RGB) images with 8 bit algorithms: each color plane can be described as a simple 8 bit image (hPitch = 3, the address distances are 1).
• Some algorithms can make use of the pitch values: for example if the hPitch is 1, it means that the pixels are continous in memory and the algorithm can use more effective operations.