using System; using System.Collections.Generic; using System.Drawing; using System.Drawing.Drawing2D; using System.Drawing.Imaging; using System.Text.RegularExpressions; namespace ImageProcessor.Processors { public abstract class ResizeBase : IGraphicsProcessor { public abstract Regex RegexPattern { get; } public abstract dynamic DynamicParameter { get; set; } public abstract int SortOrder { get; protected set; } public abstract Dictionary Settings { get; set; } public abstract int MatchRegexIndex(string queryString); public abstract Image ProcessImage(ImageFactory factory); protected Image ResizeImage(ImageFactory factory, int width, int height, int defaultMaxWidth, int defaultMaxHeight) { Bitmap newImage = null; Image image = factory.Image; try { int sourceWidth = image.Width; int sourceHeight = image.Height; int maxWidth = defaultMaxWidth > 0 ? defaultMaxWidth : int.MaxValue; int maxHeight = defaultMaxHeight > 0 ? defaultMaxHeight : int.MaxValue; // If height or width is not passed we assume that the standard ratio is to be kept. if (height == 0) { // Bit of simple fractional maths here. float percentWidth = Math.Abs(width/(float) sourceWidth); height = (int) Math.Floor(sourceHeight*percentWidth); } if (width == 0) { float percentHeight = Math.Abs(height/(float) sourceHeight); width = (int) Math.Floor(sourceWidth*percentHeight); } if (width > 0 && height > 0 && width <= maxWidth && height <= maxHeight) { // Dont use an object initializer here. newImage = new Bitmap(width, height, PixelFormat.Format32bppPArgb); newImage.Tag = image.Tag; using (Graphics graphics = Graphics.FromImage(newImage)) { // We want to use two different blending algorithms for enlargement/shrinking. // Bicubic is better enlarging for whilst Bilinear is better for shrinking. // http://www.codinghorror.com/blog/2007/07/better-image-resizing.html if (image.Width < width && image.Height < height) { // We are making it larger. graphics.SmoothingMode = SmoothingMode.AntiAlias; graphics.InterpolationMode = InterpolationMode.HighQualityBicubic; graphics.PixelOffsetMode = PixelOffsetMode.HighQuality; graphics.CompositingQuality = CompositingQuality.HighQuality; } else { // We are making it smaller. graphics.SmoothingMode = SmoothingMode.None; // Contrary to everything I have read bicubic is producing the best results. graphics.InterpolationMode = InterpolationMode.HighQualityBicubic; graphics.PixelOffsetMode = PixelOffsetMode.None; graphics.CompositingQuality = CompositingQuality.HighSpeed; } // An unwanted border appears when using InterpolationMode.HighQualityBicubic to resize the image // as the algorithm appears to be pulling averaging detail from surFlooring pixels beyond the edge // of the image. Using the ImageAttributes class to specify that the pixels beyond are simply mirror // images of the pixels within solves this problem. using (ImageAttributes wrapMode = new ImageAttributes()) { wrapMode.SetWrapMode(WrapMode.TileFlipXY); Rectangle destRect = new Rectangle(0, 0, width, height); graphics.DrawImage(image, destRect, 0, 0, sourceWidth, sourceHeight, GraphicsUnit.Pixel, wrapMode); } // Reassign the image. image.Dispose(); image = newImage; } } } catch { if (newImage != null) { newImage.Dispose(); } } return image; } } }