Building the Right Environment to Support AI, Machine Learning and Deep Learning
Unlike traditional rotation of images, where every n'th pixel is sampled and copied to the result image, this template provides much more accurate image rotation features (weighing the pixels).
The template is instantiated with a specific pixel format and must be inherited to support pixel access functions for a specific bitmap implementation. This provides abstaction of the rotation mechanism from the actual bitmap representation.
- Provides professional quality image rotation.
- Code is optimized for image quality, not speed.
- Rotated image retains size and aspect ratio of source image (destination image size is usally bigger).
- Supports double precision rotation angles (0..360).
- Supports generic bitmap structures. Specific bitmap structure is defined by inheritance.
- Template based - no need for libraries, DLLs etc. No linkage problems.
How to use the rotation template:You must derive you own class from the rotation template.
class CRotateByShearRGB : public CRotateByShear<COLORREF>will define a concrete class where each pixel is of COLORREF type.
Next, you must override 5 abstract functions which support creation, destruction of bitmap buffers and pixel level access to the bitmap buffers.
Next, you instantiate your derived class and call the AllocAndRotate(...) function which allocates a new destination bitmap buffer and rotates the input image into it.
For further instructions please view the attached sample command-line project (Rotate.cpp).
It is a simple command line utility which reads PPM (ASCII or binary) image file into a DIB structure, rotates it and saves it to another PPM image file.