cs50 BLUR PSET4 invalid or unsupported image format - c

Question

This is a problem form cs50 problem set 4 (less comfortable). I'm currently stuck on the blur part. It compiles fine but the output is "invalid or unsupported image format".
RGBTRIPLE temp[height][width];
for (int i = 0; i < height; i++)
{
for (int j = 0; j < width; j++)
{
float red_sum = 0;
float blue_sum = 0;
float green_sum = 0;
int count = 0;
for (int x = i-1 ; x <= i+1 ; i++)
{
for (int y = j-1 ; y <= j+1 ; j++)
{
if ( x >= 0 && x < height && y >= 0 && y < width )
{
red_sum += image[x][y].rgbtRed;
blue_sum += image[x][y].rgbtBlue;
green_sum += image[x][y].rgbtGreen;
count ++;
}
}
}
int red_avg = round(red_sum/count);
int blue_avg = round(blue_sum/count);
int green_avg = round(green_sum/count);
temp[i][j].rgbtRed = red_avg;
temp[i][j].rgbtBlue = blue_avg;
temp[i][j].rgbtGreen = green_avg;
}
}
for (int i = 0; i < height; i++)
{
for (int j = 0; j < width; j++)
{
image[i][j]=temp[i][j];
}
}

Answer 1

You have 2 typos which cause that you are corrupting lots of memory
for (int i = 0; i < height; i++)
{
for (int j = 0; j < width; j++)
{
float red_sum = 0;
float blue_sum = 0;
float green_sum = 0;
int count = 0;
for (int x = i-1 ; x <= i+1 ; i++) // << This loops until i overflows
{
for (int y = j-1 ; y <= j+1 ; j++) // << this loop until j overflows
{
// in second iteration you will already have negative values but due to these checks you don't access `image` out of bounds.
if ( x >= 0 && x < height && y >= 0 && y < width )
{
red_sum += image[x][y].rgbtRed;
blue_sum += image[x][y].rgbtBlue;
green_sum += image[x][y].rgbtGreen;
count ++;
}
}
}
int red_avg = round(red_sum/count);
int blue_avg = round(blue_sum/count);
int green_avg = round(green_sum/count);
// After processing data for first pixel, both `i`and `j` contain negative values causing out of bounds accesses.
temp[i][j].rgbtRed = red_avg;
temp[i][j].rgbtBlue = blue_avg;
temp[i][j].rgbtGreen = green_avg;
}
}
Due to overflow and out of bounds accesses you have lots of undefined behaviour in your program.
You can corrupt any kind of memory including your own counter variables, the image header or anything else.
Anything can happen...
The statement about negative values assumes that two's complement is used and integers overflow from highest positive value to lowest negative value. Generally this is implementation defined/undefined.
I only wonder why file type error is given because I would assume that is done before your function is called. But maybe CS50 does it again afterwards.