nQuant.cs

KO | EN

gitlite — search

#python #android #java #tensorflow #javascript #machine-learning #react #kafka #pytorch #iot #rest-api #linux

★ 24 Open GitHub ↗

Fast pairwise nearest neighbor based algorithm with C# console

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# nQuant.cs Color Quantizer
Fast pairwise nearest neighbor based algorithm with C# console

nQuant.cs is a C# color quantizer producing high quality 256 color 8 bit PNG images using an algorithm optimized for the highest quality possible.

Another advantage of nQuant.cs is that it is a .net library that you can integrate nicely with your own C# code while many of the popular quantizers only provide command line implementations. nQuant.cs also provides a command line wrapper in case you want to use it from the command line.

Less artifacts by using advanced dithering techniques such as Generalized Hilbert ("gilbert") space-filling curve and partial Blue noise distribution to diffuse the minimized quantization errors.

If you are using C#, you would call nQuant as follows:

```cs
    bool dither = true;
    var quantizer = new PnnQuant.PnnQuantizer();
    using(var bitmap = new Bitmap(sourcePath))
    {
        try
        {                    
            using (var dest = quantizer.QuantizeImage(bitmap, pixelFormat, maxColors, dither))
            {
                dest.Save(targetPath, ImageFormat.Png);
                System.Console.WriteLine("Converted image: " + Path.GetFullPath(targetPath));
            }
        }
        catch (Exception q)
        {
            System.Console.WriteLine(q.StackTrace);
        }
    }
```

More importantly, a parallel genetic algorithm called PNNLAB+ is proposed for converting a sequence of similar images under the same palette. 
```cs
 var alg = new APNsgaIII<PnnLABGAQuantizer>(new PnnLABGAQuantizer(new PnnLABQuantizer(), bitmaps, maxColors));
 alg.Run(999, -Double.Epsilon);
 using (var pGAq = alg.Result) {
 System.Console.WriteLine("\n" + pGAq.Result);
 var imgs = pGAq.QuantizeImage(dither);
 for (int i = 0; i < imgs.Count; ++i) {
 var fname = Path.GetFileNameWithoutExtension(paths[i]); 
 var destPath = Path.Combine(targetPath, fname) + " - PNNLAB+quant" + maxColors + ".png";
 imgs[i].Save(destPath, ImageFormat.Png);
 System.Console.WriteLine("Converted image: " + Path.GetFullPath(destPath));
 }					
 }
```

OTSU method (OTSU) is a global adaptive binarization threshold image segmentation algorithm. This algorithm takes the maximum inter class variance between the background and the target image as the threshold selection rule.
```cs
 var quantizer = new OtsuThreshold.Otsu();
 using(var bitmap = new Bitmap(sourcePath))
 {
 try
 { 
 using (var dest = quantizer.ConvertGrayScaleToBinary(bitmap))
 {
 dest.Save(targetPath, ImageFormat.Png);
 System.Console.WriteLine("Converted black and white image: " + Path.GetFullPath(targetPath));
 }
 }
 catch (Exception q)
 {
 System.Console.WriteLine(q.StackTrace);
 }
 }
```
Example image: <img src="https://user-images.githubusercontent.com/26831069/142559831-f8f6f2ce-487e-4353-8aa1-7845706e7833.png" />
Resulted image: <pre><img src="https://user-images.githubusercontent.com/26831069/142559920-88143e07-2787-46a2-a07c-cccf5a39065a.png" /></pre>

If you are using the command line. Assuming you are in the same directory as nQuant.exe and nQuant.Master.dll, you would enter:
`nQuant yourImage.jpg /o yourNewImage.png`

To switch algorithms, `/a otsu` can perform the above black and white conversion.

nQuant will quantize yourImage.jpg and create yourNewImage.png in the same directory.

There are a few configuration arguments you can optionaly use to try and influence how the image gets quantized. These are explained in the console application.

nQuant.cs

Explore Similar Repositories

Was this content helpful?

Recent Feedback

Download README