抱歉，我没有完全理解您的算法。它似乎表现得像
最近邻域采样，但变量diff的使用
对我来说看起来很模糊。
处理图像的常用方法是存储整个像素
最初进入记忆。
然后您可以随机访问以下任何像素
流程和代码会更直接。
顺便说一句，您发布的输入图像太小而无法识别问题，因为
很细的线条和小点很容易折叠，即使过滤
如果算法合适。
建议使用较大的图片进行评估。
这是我根据你的代码重写的：
#include <stdio.h>
#include <stdlib.h>
#include "bmp.h"
int main(int argc, char *argv[])
{
// ensure proper usage
if (argc != 4) {
fprintf(stderr, "Usage: resize n infile outfile\n");
return 1;
}
// read the scaling factor
float f = atof(argv[1]);
if(f <= 0 || f > 1) {
fprintf(stderr, "f, the resize factor, must be between 0 and 1.\n");
return 1;
}
char *infile = argv[2];
char *outfile = argv[3];
// open input file
FILE *inptr = fopen(infile, "r");
if (inptr == NULL) {
fprintf(stderr, "Could not open %s.\n", infile);
return 2;
}
// open output file
FILE *outptr = fopen(outfile, "w");
if (outptr == NULL) {
fclose(inptr);
fprintf(stderr, "Could not create %s.\n", outfile);
return 3;
}
// read infile's BITMAPFILEHEADER
BITMAPFILEHEADER bf;
fread(&bf, sizeof(BITMAPFILEHEADER), 1, inptr);
// read infile's BITMAPINFOHEADER
BITMAPINFOHEADER bi;
fread(&bi, sizeof(BITMAPINFOHEADER), 1, inptr);
// ensure infile is (likely) a 24-bit uncompressed BMP 4.0
if (bf.bfType != 0x4d42 || bf.bfOffBits != 54 || bi.biSize != 40 ||
bi.biBitCount != 24 || bi.biCompression != 0) {
fclose(outptr);
fclose(inptr);
fprintf(stderr, "Unsupported file format.\n");
return 4;
}
BITMAPFILEHEADER bf_resize = bf;
BITMAPINFOHEADER bi_resize = bi;
bi_resize.biWidth = bi.biWidth * f;
bi_resize.biHeight = bi.biHeight * f;
int padding = bi.biWidth % 4; // you can simplify the calculation
int padding_resize = bi_resize.biWidth % 4;
bi_resize.biSizeImage = (bi_resize.biWidth * sizeof(RGBTRIPLE) + padding_resize) * bi_resize.biHeight;
bf_resize.bfSize = bi_resize.biSizeImage + sizeof(BITMAPFILEHEADER) + sizeof(BITMAPINFOHEADER);
// allocate mamory for the rgb triplets of the original (input) image
RGBTRIPLE *pix = malloc(sizeof(RGBTRIPLE) * bi.biWidth * bi.biHeight);
if (pix == NULL) {
fprintf(stderr, "malloc failed.\n");
return 5;
}
// temporary storage
RGBTRIPLE triple;
// read the entire pixels of the original image and store into the memory
for (int i = 0; i < bi.biHeight; i++) {
for (int j = 0; j < bi.biWidth; j++) {
fread(&triple, sizeof(RGBTRIPLE), 1, inptr);
pix[i * bi.biWidth + j] = triple;
}
// skip over padding, if any
fseek(inptr, padding, SEEK_CUR);
}
// write outfile's header
fwrite(&bf_resize, sizeof(BITMAPFILEHEADER), 1, outptr);
fwrite(&bi_resize, sizeof(BITMAPINFOHEADER), 1, outptr);
// write the pixels of destination (resized) image
for (int i = 0; i < bi_resize.biHeight; i++) {
for (int j = 0; j < bi_resize.biWidth; j++) {
// calculate the corresponding coorinates in the original image
int m = (i / f + 0.5); // +0.5 for rounding
if (m > bi.biHeight - 1) { // limit the value
m = bi.biHeight - 1;
}
int n = (j / f + 0.5);
if (n > bi.biWidth - 1) {
n = bi.biWidth - 1;
}
// pick the pixel value at the coordinate
triple = pix[m * bi.biWidth + n];
// write RGB triplet to outfile
fwrite(&triple, sizeof(RGBTRIPLE), 1, outptr);
}
// padding for the output image, if any
for (int j = 0; j < padding_resize; j++) {
fputc(0x00, outptr);
}
}
free(pix);
fclose(inptr);
fclose(outptr);
return 0;
}
输入图像：
f=0.5 的输出图像：