Finding out the ContentType of a Image from the byte[]

Question

Is there a way to find out what the ContentType of an image is from only the original bytes?

At the moment I have a database column that stores only the byte[], which I use to display an image on a web page.

MemoryStream ms = new MemoryStream(imageBytes);
Image image = Image.FromStream(ms);
image.Save(context.HttpContext.Response.OutputStream, <--ContentType-->);

I could of course just save the ContentType in another column in the table, but just wondered if there was another way e.g. maybe .Net has a way to interrogate the data to get the type.

score 16 · Accepted Answer · answered Aug 07 '09 at 15:57

16

Check out this file signatures table.

answered Aug 07 '09 at 15:57

vit

2,675
2
18
15

Cheers for the link. I did some searching, now I know what to look for and this does appear the way to go. – Nick Clarke Aug 07 '09 at 16:32
I have put a working version of the code below, based on your idea. – Nick Clarke Aug 07 '09 at 17:32

score 15 · Answer 2 · edited May 23 '17 at 11:54

File/magic signatures was the way to go. Below is the working version of the code.

Ref: Stackoverflow - Getting image dimensions without reading the entire file

ImageFormat contentType = ImageHelper.GetContentType(this.imageBytes);

MemoryStream ms = new MemoryStream(this.imageBytes);
Image image = Image.FromStream(ms);
image.Save(context.HttpContext.Response.OutputStream, contentType);

And then the helper class:

public static class ImageHelper
{
    public static ImageFormat GetContentType(byte[] imageBytes)
    {
        MemoryStream ms = new MemoryStream(imageBytes);

        using (BinaryReader br = new BinaryReader(ms))
        {
            int maxMagicBytesLength = imageFormatDecoders.Keys.OrderByDescending(x => x.Length).First().Length;

            byte[] magicBytes = new byte[maxMagicBytesLength];

            for (int i = 0; i < maxMagicBytesLength; i += 1)
            {
                magicBytes[i] = br.ReadByte();

                foreach (var kvPair in imageFormatDecoders)
                {
                    if (magicBytes.StartsWith(kvPair.Key))
                    {
                        return kvPair.Value;
                    }
                }
            }

            throw new ArgumentException("Could not recognise image format", "binaryReader");
        }
    }

    private static bool StartsWith(this byte[] thisBytes, byte[] thatBytes)
    {
        for (int i = 0; i < thatBytes.Length; i += 1)
        {
            if (thisBytes[i] != thatBytes[i])
            {
                return false;
            }
        }
        return true;
    }

    private static Dictionary<byte[], ImageFormat> imageFormatDecoders = new Dictionary<byte[], ImageFormat>()
    {
        { new byte[]{ 0x42, 0x4D }, ImageFormat.Bmp},
        { new byte[]{ 0x47, 0x49, 0x46, 0x38, 0x37, 0x61 }, ImageFormat.Gif },
        { new byte[]{ 0x47, 0x49, 0x46, 0x38, 0x39, 0x61 }, ImageFormat.Gif },
        { new byte[]{ 0x89, 0x50, 0x4E, 0x47, 0x0D, 0x0A, 0x1A, 0x0A }, ImageFormat.Png },
        { new byte[]{ 0xff, 0xd8 }, ImageFormat.Jpeg },
    };

Out of curiosity, is there a reason your loops increment with "i += 1" instead of "i++"? — Portman, Aug 11 '09 at 20:20
I adapted the code from another Stackoverflow question/answer (mentioned above). The code has since changed as I also understand i++ more than i +=1. — Nick Clarke, Aug 12 '09 at 09:16
@Portman, personal styles and preferences. Some people think `i++` is clearer, others think that it is less "best practice". See this: http://stackoverflow.com/questions/971312/why-avoid-increment-and-decrement-operators-in-javascript — ANeves, Oct 25 '12 at 12:06

Joost Aarts · Answer 3 · 2012-01-09T00:16:41.790

This worked for me, ms being the memorystream. The downside is that it has to load the image.

Dim fmt As System.Drawing.Imaging.ImageFormat
Dim content As String

Using bmp As New Drawing.Bitmap(ms)
    fmt = bmp.RawFormat
End Using

Select Case fmt.Guid
    Case Drawing.Imaging.ImageFormat.Bmp.Guid
        content = "image/x-ms-bmp"

    Case Drawing.Imaging.ImageFormat.Jpeg.Guid
        content = "image/jpeg"

    Case Drawing.Imaging.ImageFormat.Gif.Guid
        content = "image/gif"

    Case Drawing.Imaging.ImageFormat.Png.Guid
        content = "image/png"

    Case Else
        content = "application/octet-stream"

End Select

score 4 · Answer 4 · answered Jan 08 '16 at 12:54

Rewrote the method of Nick Clarke a bit using Linq:

public class Program
{
   public static void Main(string[] args)
   {
      byte[] byteArray = File.ReadAllBytes(@"C:/users/Alexander/image.jpg");
      ImagePartType type = byteArray.GetImageType();
   }
}


public static class ImageHelper
{
    public static ImagePartType GetImageType(this byte[] imageBytes)
    {
        foreach(var imageType in imageFormatDecoders)
        {
            if (imageType.Key.SequenceEqual(imageBytes.Take(imageType.Key.Length)))
                return imageType.Value;
        }

        throw new ArgumentException("Imagetype is unknown!");
    }

    private static Dictionary<byte[], ImagePartType> imageFormatDecoders 
                     = new Dictionary<byte[], ImagePartType>()
    {
       { new byte[]{ 0x42, 0x4D }, ImagePartType.Bmp},
       { new byte[]{ 0x47, 0x49, 0x46, 0x38, 0x37, 0x61 }, ImagePartType.Gif },
       { new byte[]{ 0x47, 0x49, 0x46, 0x38, 0x39, 0x61 }, ImagePartType.Gif },
       { new byte[]{ 0x89, 0x50, 0x4E, 0x47, 0x0D, 0x0A, 0x1A, 0x0A }, ImagePartType.Png },
       { new byte[]{ 0xff, 0xd8 }, ImagePartType.Jpeg }
    };
}

I used ImagePartType because I'm working with Open XML SDK at the moment, but just change the type in the dictionary :)

score 1 · Answer 5 · answered Aug 07 '09 at 15:58

1

There's no standard way to detect content type from a stream built-in .NET. You could implement your own algorithm that could achieve this for some well-known image formats by reading the first few bytes and trying to match the format.

answered Aug 07 '09 at 15:58

Darin Dimitrov

1,023,142
271
3,287
2,928

score 0 · Answer 6 · answered Aug 07 '09 at 15:58

0

Does the RawFormat property work for you?

MemoryStream ms = new MemoryStream(imageBytes);
Image image = Image.FromStream(ms);
image.Save(context.HttpContext.Response.OutputStream, image.RawFormat);

answered Aug 07 '09 at 15:58

Jon Skeet

1,421,763
867
9,128
9,194

RawFormat seems to work for jpeg and gif but not png. For a test I also tried hardcoding it to jpeg and this worked the same (for all and not png), I guess .Net applies more logic to png. – Nick Clarke Aug 07 '09 at 16:11
From other examples I have seen RawFormat appears to only work when the image was loaded from the filesystem and not created via bytes (db column). – Nick Clarke Aug 07 '09 at 17:40

Finding out the ContentType of a Image from the byte[]

6 Answers6