Allow loading 8bpp and lower color images without dithering the image and changing the palette

[a740g]

QB64-PE uses stb_image & dr_pcx libraries to load all image files when requests are made using _LOADIMAGE.

Both stb_image and dr_pcx automatically de-palettizes the image on load. So, there is no way to tell what kind of color information these images had after the image data is received from these libraries.

See nothings/stb#510 and nothings/stb#1124

So when the user loads a 24-bit image using _LOADIMAGE("my_image.jpg", 256), then:

1. The image is converted to 32bpp by stb_image while loading
2. Once we have the 32bpp data, _LOADIMAGE converts the image to 8bpp and applies ordered dithering
3. An adaptive palette is created or the image is remapped to QB64-PE master palette (depending on the _LOADIMAGE bpp parameter)

This results in an acceptable 8-bit image quality.

However, when the user loads an 8-bit palettized image using _LOADIMAGE("my_image.gif", 256), stb_image de-palettizes the image and the same steps are applied to the data (as noted above). Hence, there are two side-effects:

1. The color palette is not the same compared to the original image it was loaded from
2. Some artifacts may be seen in the final image due to the ordered dithering step