这些年来我多次遇到这个问题，并且仍然希望有一种简单的方法可以做到这一点，但我已经错过了。 我经常使用条形码。它们通常由白色背景上的黑色点或线条组成。当边缘清晰且线条或点的尺寸精确时，条形码读取器通常工作得更快、更准确。
大多数条形码生成算法都会为您提供紧凑的条形码，通常最小元素大小为一个像素。典型的 QR 码可以容纳在 21 x 21 的网格中。如果在大多数打印机上逐像素打印，则该尺寸太小而无法查看，并且通常会按比例放大。放大的结果取决于所使用的方法，尽管有时您可以选择，但通常您没有选择使图像合适。即使直接打印也常常会出现预期的灰色伪影或某种形式的抖动。我发现最一致的方法是在每天在其他地方使用图像之前对其进行缩放，例如 Microsoft Word、lightburn 和我使用的其他一些仍然让我头疼的地方。
下面我将详细介绍我的尝试并展示结果。我将其限制为位图，因为我当前的项目不需要在这里使用矢量。
我当前的最佳分辨率并不漂亮，速度很慢，尽管我可以通过锁定位图中的位来提高速度，但我希望有人能给出一个非常简单的答案，而我这次再次完全错过了搜索。
这是在 GIMP 中放大的简单 QR 码图像。
问题是，如果按比例放大，它通常会看起来像这样:
下面我创建了一个小测试程序来遍历我所知道的所有不同模式，然后生成我在下面复制的图像矩阵。我当前使用的版本是模式 99，它涉及检查每个像素并绘制一个正方形。
大家有更好的想法吗？
using System;
using System.Collections.Generic;
using System.Drawing;
using System.IO;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
using System.Windows;
using System.Windows.Media.Imaging;
using System.Drawing.Drawing2D;
using System.Drawing.Imaging;
namespace PSWpfCommon.Images
{
public partial class WPFImageHelper
{
public class Test
{
public int smode = 0;
public int imode = 0;
public int mode = 0;
public string title = "";
public Test(int s, int i, int m, string t)
{
smode = s;
imode = i;
mode = m;
title = t;
}
}
public static Bitmap TestImage()
{
byte[] img =
{
0x89, 0x50, 0x4E, 0x47, 0x0D, 0x0A, 0x1A, 0x0A, 0x00, 0x00, 0x00, 0x0D, 0x49, 0x48, 0x44,
0x52, 0x00, 0x00, 0x00, 0x15, 0x00, 0x00, 0x00, 0x15, 0x08, 0x06, 0x00, 0x00, 0x00, 0xA9,
0x17, 0xA5, 0x96, 0x00, 0x00, 0x00, 0xCF, 0x49, 0x44, 0x41, 0x54, 0x38, 0xCB, 0x9D, 0x54,
0x5B, 0x0E, 0xC3, 0x30, 0x08, 0xB3, 0xAB, 0xDC, 0xFF, 0xCA, 0xDE, 0xC7, 0xD4, 0x88, 0x7A,
0x3C, 0xD2, 0x21, 0x55, 0xAD, 0x48, 0xA0, 0xC6, 0x80, 0x09, 0x40, 0x28, 0x4C, 0x12, 0x48,
0xEE, 0x6F, 0x00, 0x20, 0xF9, 0xF0, 0x67, 0xB6, 0x62, 0x40, 0xB4, 0x98, 0xAC, 0x4A, 0x50,
0xC5, 0x2D, 0x4F, 0xE2, 0x97, 0x23, 0xB2, 0xEE, 0xE7, 0x31, 0xEE, 0xC2, 0xA1, 0x75, 0x89,
0xD3, 0xF2, 0x27, 0x73, 0x5E, 0x8F, 0x93, 0x56, 0x01, 0x53, 0xA2, 0xEC, 0x7C, 0x39, 0x2F,
0x19, 0xCA, 0x58, 0x7A, 0xA4, 0xA0, 0x8A, 0xA3, 0x0E, 0x6A, 0x7A, 0x5B, 0xFE, 0x45, 0x12,
0xF1, 0xF1, 0x44, 0x59, 0x73, 0xFC, 0x5E, 0x8C, 0x25, 0xF9, 0xED, 0xBE, 0xA4, 0x47, 0x49,
0xDD, 0x18, 0x79, 0xF9, 0x25, 0xA7, 0xD9, 0xA6, 0x74, 0xE3, 0xE3, 0x48, 0x9D, 0xE3, 0x55,
0xA1, 0x98, 0xB8, 0xCC, 0x16, 0xE4, 0xF6, 0x5F, 0xBE, 0xDF, 0x8E, 0x74, 0x42, 0x9F, 0x4D,
0xC0, 0x0F, 0xA7, 0xFE, 0x76, 0x14, 0x5D, 0x65, 0xDB, 0x3F, 0xA9, 0x54, 0xC7, 0x9D, 0x8B,
0xCD, 0x7D, 0x3E, 0xAA, 0xD4, 0x24, 0x77, 0x99, 0x7F, 0x54, 0xA9, 0xAA, 0x69, 0x7E, 0x3F,
0xCE, 0xEA, 0xFA, 0x67, 0x9B, 0x3A, 0x2A, 0x24, 0x9D, 0x49, 0x9F, 0x23, 0x99, 0x64, 0x71,
0x9D, 0xA8, 0x51, 0xC5, 0x6F, 0x36, 0x21, 0x5B, 0xF9, 0x3B, 0x95, 0xAA, 0x1A, 0x52, 0xAD,
0xB1, 0x24, 0x7C, 0x00, 0x22, 0x8E, 0xDE, 0x4C, 0xC4, 0x8F, 0x11, 0x7F, 0x00, 0x00, 0x00,
0x00, 0x49, 0x45, 0x4E, 0x44, 0xAE, 0x42, 0x60, 0x82,
};
Bitmap qr;
using (var ms = new MemoryStream(img))
{
qr = new Bitmap(ms);
}
int factor = 4;
var l = new List<Test>();
for (int i = 0; i <= 7; i++)
for (int s = 0; s <= 4; s++)
l.Add(new Test(s, i, 0, $"s={s} i={i}"));
l.Add(new Test(2, 8, 99, $"Mode 99"));
Bitmap fullimage = new Bitmap(((qr.Width * factor) + 20) * 5, ((qr.Height * factor) + 20) * 9);
fullimage.SetResolution(72, 72);
var font = new Font("Arial", 10);
using (var grPhoto = Graphics.FromImage(fullimage))
using (var blackbrush = new SolidBrush(System.Drawing.Color.Black))
using (var whitebrush = new SolidBrush(System.Drawing.Color.White))
{
grPhoto.InterpolationMode = InterpolationMode.High;
grPhoto.FillRectangle(whitebrush, 0, 0, fullimage.Width, fullimage.Height);
foreach (var t in l)
{
var newqr = GrowImage(qr, factor: 4, mode: t.mode, imode: t.imode, smode: t.smode);
grPhoto.DrawImage(newqr,
t.smode * ((qr.Width * factor) + 20),
t.imode * ((qr.Height * factor) + 20));
grPhoto.DrawString(t.title, font, blackbrush,
t.smode * ((qr.Width * factor) + 20),
(t.imode + 1) * ((qr.Height * factor) + 20) - 20);
}
}
fullimage.Save(@"c:\temp\newqr.png", ImageFormat.Png);
return null;
}
public static Bitmap GrowImage(Bitmap im, int factor = 4, int mode = 1, int imode = 0, int smode = 0, int border = 2)
{
bool translate = true;
var bmPhoto = new Bitmap(im.Width * factor + 2 * border, im.Height * factor + border * 2, PixelFormat.Format24bppRgb);
bmPhoto.SetResolution(72, 72);
using (var grPhoto = Graphics.FromImage(bmPhoto))
using (var blackbrush = new SolidBrush(System.Drawing.Color.Black))
using (var whitebrush = new SolidBrush(System.Drawing.Color.White))
{
grPhoto.FillRectangle(whitebrush, 0, 0, bmPhoto.Width, bmPhoto.Height);
switch (smode)
{
case 0:
grPhoto.SmoothingMode = SmoothingMode.Default;
break;
case 1:
grPhoto.SmoothingMode = SmoothingMode.AntiAlias;
break;
case 2:
grPhoto.SmoothingMode = SmoothingMode.HighQuality;
break;
case 3:
grPhoto.SmoothingMode = SmoothingMode.HighSpeed;
break;
case 4:
grPhoto.SmoothingMode = SmoothingMode.None;
break;
default:
break;
}
switch (imode)
{
case 0:
grPhoto.InterpolationMode = InterpolationMode.Default;
break;
case 1:
grPhoto.InterpolationMode = InterpolationMode.Bicubic;
break;
case 2:
grPhoto.InterpolationMode = InterpolationMode.Bilinear;
break;
case 3:
grPhoto.InterpolationMode = InterpolationMode.High;
break;
case 4:
grPhoto.InterpolationMode = InterpolationMode.HighQualityBicubic;
break;
case 5:
grPhoto.InterpolationMode = InterpolationMode.HighQualityBilinear;
break;
case 6:
grPhoto.InterpolationMode = InterpolationMode.Low;
break;
case 7:
grPhoto.InterpolationMode = InterpolationMode.NearestNeighbor;
break;
default:
break;
}
switch (mode)
{
case 99:
// These are what worked best for me...
grPhoto.SmoothingMode = SmoothingMode.None;
grPhoto.InterpolationMode = System.Drawing.Drawing2D.InterpolationMode.NearestNeighbor;
for (int x = 0; x < im.Width; x++)
for (int y = 0; y < im.Height; y++)
{
var g = im.GetPixel(x, y);
if (g.R < 120 && g.B < 120) // Just being really lazy here... if the pixel has not much blue and not much red I'll treat it as black
{
grPhoto.FillRectangle(blackbrush, border + factor * x, border + factor * y, factor, factor);
}
}
translate = false;
break;
default:
break;
}
if (translate) // If we used mode 99, don't draw the image
grPhoto.DrawImage(im, new System.Drawing.Rectangle(border, border, im.Width * factor, im.Height * factor), 0, 0, im.Width, im.Height, System.Drawing.GraphicsUnit.Pixel);
}
return bmPhoto;
}
}
}
最佳答案
这就是我如何使用 DrawImage 和最近邻采样一次性调整 Bitmap 的大小。
在您的情况下，我会根据您的图案的几何尺寸(例如，21 x 21)准备每个逻辑点一个像素的原始位图，然后我会将该位图的大小调整为可打印/可显示的输出格式使用我在下面提供的代码使用 Interpolationmode.NearestNeighbor 进行采样。
设置PixelOffsetMode.HighQuality非常重要。 (除非你想将坐标调整-0.5f，-0.5f，这很烦人，所以不要打扰)。
public static Bitmap Resize(this Bitmap oldBitmap, int newWidth, int newHeight)
{
Bitmap newBitmap = new Bitmap(newWidth, newHeight, oldBitmap.PixelFormat);
using (Graphics g = Graphics.FromImage(newBitmap)) {
RectangleF dst = new RectangleF(0, 0, newWidth, newHeight);
RectangleF src = new RectangleF(0, 0, oldBitmap.Width, oldBitmap.Height);
g.InterpolationMode = System.Drawing.Drawing2D.InterpolationMode.NearestNeighbor;
g.PixelOffsetMode = System.Drawing.Drawing2D.PixelOffsetMode.HighQuality;
g.DrawImage(oldBitmap, dst, src, GraphicsUnit.Pixel);
}
return newBitmap;
}
关于c# - 在 C# 中放大条形码图像的最佳无损方法是什么，我们在Stack Overflow上找到一个类似的问题： https://stackoverflow.com/questions/71055395/