public static void GetSmallestBitmap(Bitmap sourceImage, ScanBitDepth bitDepth, bool highQuality, out Bitmap bitmap, out MemoryStream encodedBitmap, out ImageFormat imageFormat)
{
// Defaults for out arguments
bitmap = null;
encodedBitmap = null;
imageFormat = ImageFormat.Png;
// Store the image in as little space as possible
if (bitDepth == ScanBitDepth.BlackWhite)
{
// Store as a 1-bit bitmap
// This is lossless and takes up minimal storage (best of both worlds), so highQuality is irrelevant
bitmap = (Bitmap)BitmapHelper.CopyToBpp(sourceImage, 1).Clone();
// Note that if a black and white image comes from native WIA, bitDepth is unknown,
// so the image will be png-encoded below instead of using a 1-bit bitmap
}
else if (highQuality)
{
// Store as PNG
// Lossless, but some images (color/grayscale) take up lots of storage
encodedBitmap = EncodeBitmap(sourceImage, ImageFormat.Png);
}
else
{
// Store as PNG/JPEG depending on which is smaller
var pngEncoded = EncodeBitmap(sourceImage, ImageFormat.Png);
var jpegEncoded = EncodeBitmap(sourceImage, ImageFormat.Jpeg);
if (pngEncoded.Length <= jpegEncoded.Length)
{
// Probably a black and white image (from native WIA, so bitDepth is unknown), which PNG compresses well vs. JPEG
encodedBitmap = pngEncoded;
jpegEncoded.Dispose();
}
else
{
// Probably a color or grayscale image, which JPEG compresses well vs. PNG
encodedBitmap = jpegEncoded;
pngEncoded.Dispose();
imageFormat = ImageFormat.Jpeg;
}
}
}
