|
You don't mention the format of your image. In some formats the color data is in each pixel, in others the pixels are an index into a color table. It makes a difference as to which part you search. If the color is in the pixels, unless your image is artificially generated or size of the pixels in color value is small, you're not likely to get many exact matches.
If you don't have the data, you're just another a**hole with an opinion.
|
|
|
|
|
Hi,
I need to write a vc++ program to find the Codec Name of any DirecShow suppoting format video. How can i do that using directshow interfaces and apis?
Apart from DirectShow is there any other api of library which we can use for that.
Thanks in advance.
Kuldeep
|
|
|
|
|
Hi all,
i make one application in which i read the image file of type *.lbl using Bitmap.FromFile(filepath) at that time it will give me an error of "Out of memory".
I know Bitmap supports only BMP ,GIF,JPEG ,PNG ,TIFF. So can any one suggest me what should i do read my own extension (*.lbl) file using Bitmap or any other methode ????
Thanks in advance..
Rana Krishnraj
|
|
|
|
|
Rana Krishnraj,
Well, using the NET Framework class is clearly not going to work because it expects to see a specific, known data format.
If you have your own image file format, (and, of course, we don't know anything about it), you could:
(1) Convert it to a convenient format that your NET Framework bitmap class instance will recognize. I would suggest reading up on the native Win32 Bitmap[^] data structure, and create a valid bitmap (actually, a , DIB, Device Independent Bitmap[^]) in memory.
(2) Write a custom file reader that adequately processes your custom file format.
If you don't know anything about the Bitmap data structure, you should read: Bitmap Storage MSDN[^], or, Storing an Image MSDN[^]
As Feng Yuan says: "The good thing about a DIB is that its storage is managed by the application, so the application can directly access its color table and pixel array, but GDI would not lend a hand to you in setting up a DIB."
|
|
|
|
|
I have series of images very similar to this example image:
http://img258.imageshack.us/img258/2483/exampleep8.gif[^]
As you can see there's empty white space around the text and lines. How can I remove this empty space and crop the image? The red line shows where I want to crop it.
I'm a newbie when it comes to Drawing/GDI+, I appreciate any help.
modified on Wednesday, June 11, 2008 8:25 AM
|
|
|
|
|
Problem solved here's the code:
Bitmap b = new Bitmap(filename);
Rectangle r = new Rectangle();
int topLeft = b.Width, topRight = 0, topUpper = b.Height, topBottom = 0;
string whitePixel = "ffffffff";
for (int y = 1; y < b.Height; y++)
{
for (int x = 1; x < b.Width; x++)
{
if (b.GetPixel(x, y).Name != whitePixel)
{
if (x < topLeft)
topLeft = x;
if (x > topRight)
topRight = x;
if (y < topUpper)
topUpper = y;
if (y > topBottom)
topBottom = y;
}
}
}
r.Location = new Point(topLeft-1, topUpper-1);
r.Size = new Size((topRight - topLeft)+1, (topBottom - topUpper)+1);
Bitmap b2 = b.Clone(r, b.PixelFormat);
b2.Save(newfilename, ImageFormat.Gif);
|
|
|
|
|
Hi,
I am using Video over IP Source filter. I have built a graph.
I wanted to set this video source filters "HostName" to some IP address and connect to that IP address.
Does anyone know "How to set filter property through code without popping up property page"
Thanks 
AmolM
|
|
|
|
|
Hi, I have a program that is supposed to blur bitmaps in a Gaussian type fashion.
However, all this does is change the light greys to a yellow-ey colour.
Here is some code:
Functions:
void set_pixel(int x,int y, RGBTRIPLE colour){
image[(bmp.biHeight-1-y)*bmp.biWidth+x] = colour;
}
void get_pixel(int x, int y, RGBTRIPLE colour){
colour = image[(bmp.biHeight-1-y)*bmp.biWidth+x];
}
void avg(RGBTRIPLE colour[], RGBTRIPLE merge, int size) {
for(int i=0; i<size; i++)="" {<br="" mode="hold" /> merge.rgbtBlue += colour[i].rgbtBlue;
merge.rgbtGreen += colour[i].rgbtGreen;
merge.rgbtRed += colour[i].rgbtRed;
}
merge.rgbtBlue = merge.rgbtBlue / size;
merge.rgbtGreen= merge.rgbtGreen / size;
merge.rgbtRed = merge.rgbtRed / size;
}
Implementation:
for(int y = 0; y < bmp.biHeight; y=y+3){
for(int x = 0; x <bmp.biwidth; x="x+3){<br=""" mode="hold" /> for(int i = 0; i < 9; i++){
int yp=0, xp=0;
get_pixel(x+xp,y+yp,blank[i]);
if(xp<3)
xp++;
else{
xp = 0;
yp++;
}
}
avg(blank, Merge, 9);
set_pixel(x,y,Merge);
}}
Any help you can give would be much appreciated. Thanks
modified on Saturday, June 7, 2008 9:35 PM
|
|
|
|
|
looks messy. My image processing articles here on code project have a Gaussian blur in them.
Christian Graus
Please read this if you don't understand the answer I've given you
"also I don't think "TranslateOneToTwoBillion OneHundredAndFortySevenMillion FourHundredAndEightyThreeThousand SixHundredAndFortySeven()" is a very good choice for a function name" - SpacixOne ( offering help to someone who really needed it ) ( spaces added for the benefit of people running at < 1280x1024 )
|
|
|
|
|
You seem to have had a problem when pasting your code but unless it didn't make it, I don't see where Merge gets zeroed out between calls to avg().
If you don't have the data, you're just another a**hole with an opinion.
|
|
|
|
|
I hadn't put the zeroing of Merge in :S, thanks for pointing that out.
However, this still doesn't solve the problem.
I have also tried zeroing blank between calls of avg(). This didn't help either.
and Christian, I'm checking out your tutorials now, unfortunately I will have to translate them into C++ but I hope it will help.
EDIT: Aha!
I have just done something I realised I should have done before. I tried to create random colours for each pixel, in a noise type function without writing before blurring. I have set these up into 2 seperate projects and now the ouput is a grey screen with some black dots going in lines across the screen :S
Here's a picture of what the output is:
http://s131.photobucket.com/albums/p287/CorporateRock/blurtry.jpg[^]
"Sir, I protest. I am NOT a merry man!"
modified on Sunday, June 8, 2008 8:52 AM
|
|
|
|
|
Ah, I think I see it now. In avg(), Merge is of type RGBTRIPLE. Each color element is an unsigned char which limits its value to 0 to 255. You're adding nine elements into each color value before you take the average. On average, you've overflowed it after the second addition and wrapped around. You need to use a temporary accumulator that's at least an unsigned short but hey, go wild and make it an unsigned int.  Then you're covered if you use a bigger kernel than 3x3.
If you don't have the data, you're just another a**hole with an opinion.
|
|
|
|
|
Thanks for the advice  .
I used an unsigned int for red green and blue. This still didn't solve the problem, however as it still seems too organised and is tritone. :S
Sorry my code is so messed up.
"Sir, I protest. I am NOT a merry man!"
|
|
|
|
|
Try giving us the latest version. Kind of hard to guess how you've rattled the box.
If you don't have the data, you're just another a**hole with an opinion.
|
|
|
|
|
void set_pixel(long x,long y, RGBTRIPLE colour){
image[(bmp.biHeight-1-y)*bmp.biWidth+x] = colour;
}
void get_pixel(int x, int y, RGBTRIPLE colour){
colour = image[(bmp.biHeight-1-y)*bmp.biWidth+x];
}
void clear_pixel(RGBTRIPLE colour){
colour.rgbtBlue = 0;
colour.rgbtGreen = 0;
colour.rgbtRed = 0;
}
void avg(RGBTRIPLE colour[], RGBTRIPLE merge, int size) {
unsigned int blue = 0, green = 0, red = 0;
for(int i=0; i<size; i++)="" {<br="" mode="hold" /> blue += colour[i].rgbtBlue;
green += colour[i].rgbtGreen;
red += colour[i].rgbtRed;
}
merge.rgbtBlue = blue / size;
merge.rgbtGreen= green / size;
merge.rgbtRed = red / size;
}
<small>
int WINAPI WinMain(HINSTANCE hInstance, HINSTANCE hPrevInstance,
LPSTR lpCmdLine, int nCmdShow)
{
ZeroMemory (&bmp, sizeof bmp);
bmp.biClrImportant = 0;
bmp.biBitCount = 4;
bmp.biCompression = 0;
bmp.biPlanes = 1;
bmp.biSize = 40;
bmp.biSizeImage = 0;
bmp.biXPelsPerMeter = 0;
bmp.biYPelsPerMeter = 0;
bmp.biClrUsed = 16;
bmp.biHeight = 128;
bmp.biWidth = 128;
long paddedsize = bmp.biHeight * bmp.biWidth;
bfh.bfType = 19778;
bfh.bfReserved1 = 0;
bfh.bfReserved2 = 0;
bfh.bfOffBits = 1078;
bfh.bfSize = sizeof(BITMAPFILEHEADER) + sizeof(BITMAPINFOHEADER) + paddedsize;
hfile = CreateFile("noise.bmp",GENERIC_WRITE,FILE_SHARE_WRITE,NULL,OPEN_EXISTING,NULL,NULL);
ReadFile(hfile,&bfh,sizeof(bfh),&written,NULL);
ReadFile(hfile,&bmp,sizeof(bmp),&written,NULL);
int imagesize = bmp.biWidth*bmp.biHeight;
image = new RGBTRIPLE[imagesize];
ReadFile(hfile,image,sizeof(RGBTRIPLE),&written,NULL);</small>
RGBTRIPLE blank[9]= { 0 };
RGBTRIPLE Merge = { 0 , 0 , 0 };
long yp=0, xp=0;
for(int y = 0; y < bmp.biHeight; y=y+3){
for(int x = 0; x < bmp.biWidth; x=x+3){
for(int i = 0; i < 9; i++){
int yp=0, xp=0;
get_pixel(x+xp,y+yp,blank[i]);
if(xp<3)
xp++;
else{
xp = 0;
yp++;
} }
avg(blank, Merge, 9);
set_pixel(x,y,Merge);
for(int i = 0; i < 9; i++){
blank[i].rgbtBlue=0;
blank[i].rgbtGreen=0;
blank[i].rgbtRed=0;
}
clear_pixel(Merge);
yp=0, xp=0;
}}
EDIT: AHa! I've found something
for(int i = 0; i < 9; i++){
int yp=0, xp=0;
I declared yp and xp as int when I had already declared them as long. This was reverting them to 0 each time.
Trouble is, get_pixel() and set_pixel() give unhandled exceptions when they use anything other than x and y. I've tried "x=x+xp" but it still happens. Hmm....
I'm stumped.
"Sir, I protest. I am NOT a merry man!"
|
|
|
|
|
Ok, a few more rookie gotchas. RGBTRIPLE is a struct. In get_pixel() and clear_pixel(), you want to operate on the pixel "colour" that you pass in. However, the way you've written it, you're working on a copy so you're changing nothing. You need to declare it as a pointer or a reference, RGBTRIPLE* pColour or RGBTRIPLE& colour. If you do it as a reference, you don't need to change the code inside the functions to pointer notation. The same thing applies to merge in avg(), you're working on a copy inside the function the passed parameter doesn't change.
I also don't think you're applying the kernel to your image properly. I'm not up on gaussian off the top of my head but you need to apply it to every pixel and its immediate neighbors so bumping the index of the loops by 3 each pass doesn't do it. Usually, applying a kernel you need a source image and then build a spearate destination image, otherwise you'd be using pixels you've already changed as you progress through.
If you don't have the data, you're just another a**hole with an opinion.
|
|
|
|
|
I've changed the code as you suggested, with a source and destination file. But I get in the output file is a black image with some very dark green lines going vertically across. :
Changed it a bit to use a 1,2,1, Matrix. This gives blue and green lines. Gah this is messed up.
2,4,2,
1,2,1
Here's the entire source... again.
#include "stdafx.h"
#include <windows.h>
#include <iostream>
HANDLE hfile;
HANDLE hfile2;
DWORD written;
BITMAPFILEHEADER bfh;
BITMAPINFOHEADER bmp;
RGBTRIPLE *image;
void set_pixel(long x,long y, RGBTRIPLE& colour){
image[bmp.biHeight+y*bmp.biWidth+x] = colour;
}
void get_pixel(int x, int y, RGBTRIPLE& colour){
colour = image[bmp.biHeight+y*bmp.biWidth+x];
}
void clear_pixel(RGBTRIPLE& colour){
colour.rgbtBlue = 0;
colour.rgbtGreen = 0;
colour.rgbtRed = 0;
}
void copy_pixel(RGBTRIPLE& colour, RGBTRIPLE& colour2){
colour.rgbtBlue = colour2.rgbtBlue;
colour.rgbtGreen = colour2.rgbtGreen;
colour.rgbtRed = colour2.rgbtRed;
}
void avg(RGBTRIPLE colour[], RGBTRIPLE& merge, int size, BYTE Matrix[]) {
unsigned int blue = 0, green = 0, red = 0;
for(int i=0; i<size; i++)="" {<br="" mode="hold" /> blue += colour[i].rgbtBlue * Matrix[i] / 16;
green += colour[i].rgbtGreen * Matrix[i] / 16;
red += colour[i].rgbtRed * Matrix[i] / 16;
}
merge.rgbtBlue = blue / size;
merge.rgbtGreen= green / size;
merge.rgbtRed = red / size;
}
BYTE Matrix[9] = { 1, 2, 1,
2, 4, 2,
1, 2, 1 };
int WINAPI WinMain(HINSTANCE hInstance, HINSTANCE hPrevInstance,
LPSTR lpCmdLine, int nCmdShow)
{
ZeroMemory (&bmp, sizeof bmp);
bmp.biClrImportant = 0;
bmp.biBitCount = 8;
bmp.biCompression = 0;
bmp.biPlanes = 1;
bmp.biSize = 40;
bmp.biSizeImage = 0;
bmp.biXPelsPerMeter = 0;
bmp.biYPelsPerMeter = 0;
bmp.biClrUsed = 0;
bmp.biHeight = 128;
bmp.biWidth = 128;
long paddedsize = bmp.biHeight * bmp.biWidth;
bfh.bfType = 19778;
bfh.bfReserved1 = 0;
bfh.bfReserved2 = 0;
bfh.bfOffBits = 1078;
bfh.bfSize = sizeof(BITMAPFILEHEADER) + sizeof(BITMAPINFOHEADER) + paddedsize;
hfile = CreateFile("noisein.bmp",GENERIC_READ,FILE_SHARE_READ,NULL,OPEN_EXISTING,NULL,NULL);
hfile2 = CreateFile("noiseout.bmp",GENERIC_WRITE,FILE_SHARE_WRITE,NULL,OPEN_EXISTING,NULL,NULL);
ReadFile(hfile,&bfh,sizeof(bfh),&written,NULL);
ReadFile(hfile,&bmp,sizeof(bmp),&written,NULL);
ReadFile(hfile2,&bfh,sizeof(bfh),&written,NULL);
ReadFile(hfile2,&bmp,sizeof(bmp),&written,NULL);
int imagesize = bmp.biWidth*bmp.biHeight;
image = new RGBTRIPLE[imagesize];
ReadFile(hfile ,image,sizeof(RGBTRIPLE),&written,NULL);
ReadFile(hfile2,image,sizeof(RGBTRIPLE),&written,NULL);
RGBTRIPLE blank[9]= { 0 };
RGBTRIPLE Merge = { 0 , 0 , 0 };
int yp=0, xp=0;
for(int y = 1; y < bmp.biHeight; y++){
for(int x = 1; x < bmp.biWidth; x++){
for(int i = 0; i < 9; i++){
{
get_pixel(x+xp,y+yp,blank[i]);
if(xp<3)
xp++;
else{
xp = 0;
yp++;}
}
}
avg(blank, Merge, 9, Matrix);
set_pixel(x,y,Merge);
clear_pixel(Merge);
yp=0, xp=0;
}}
WriteFile ( hfile2, &bfh, sizeof(bfh), &bfh.bfSize, NULL );
WriteFile ( hfile2, &bmp, sizeof(bmp), &bfh.bfSize, NULL );
WriteFile ( hfile2, image, paddedsize, &bfh.bfSize, NULL );
CloseHandle(hfile);
}</iostream></windows.h>
"Sir, I protest. I am NOT a merry man!"
modified on Monday, June 9, 2008 12:00 PM
|
|
|
|
|
Ok, you have two files now but you only still have one bitmap in memory, especially the image variable. You're reading from it, operating on the pixels, and writing them back in place. You read the first file into memory and immediately read over the data in memory from the second. Not good.
If you're going to use a global for the pixel matrix, get_pixel would have to work on the original image and set_pixel on the output image. A better way would be to write general functions that have a pointer to the pixel matrix as a parameter, then you'd be set.
The other thing that has jarred my consciousness is the calcualtion of the index when you're accessing a pixel in image, ie "image[bmp.biHeight+y*bmp.biWidth+x]". Why the offset "bmp.biHeight"? Shouldn't the index to the pixel just be "y*bmp.biWidth+x"?
Hard to tell some of this since for some reason your for loops don't seem to copy well. Are you using tabs in your code rather than spaces? Anyhow, it looks like you're closing in on doing the loop ranges on row and column properly from how I remember kernels. The starting index being one, however, you need to subtract 1 from the end test or you'll run off the edge. More properly, you have to allow a border of KernelSize/2 around the edge of your image since you can't really apply a kernel to those pixels. I'm not sure the stuff in the middle is correct but it's hard to make out. You're going to force me to actually look up Gaussian blur. 
If you don't have the data, you're just another a**hole with an opinion.
|
|
|
|
|
That's the saddest face I've ever seen. I'll get right onto fixing those things. If it works, thanks for the advice.
"Sir, I protest. I am NOT a merry man!"
|
|
|
|
|
|
Ok, a few things for you to think about while I'm trying to figure out your pixel indexing scheme. You read the second file, the output file, into memory. Why? You simply want to calculate the new image and write it?
You never free the memory you allocated for the two image arrays. Won't kill you in this instance but bad practice.
It doesn't look like you close the output file before you exit. I think Windows will do this for you but I never trust Windows to do anything I can do explicitly. It saves big teeth marks in your ass if it doesn't actually do what it's advertised to do. I'll get back to you on the rest of it.
If you don't have the data, you're just another a**hole with an opinion.
|
|
|
|
|
Ok, courtesy of Wikipedia I've brushed up on Gaussian blurring.  You're trying to replace each pixel, well at last those not on an edge with a weighted average of itself and the 8 pixels surrounding it. You have your weights in the array and are multiplying the pixel by the weight and diving by the sum of the weights (16). Two things. First and most glaring is why do you divide the blurred values by "size", 9 on your case? That's going to make things dark as you just cut each pixel value to 1/9 of its value, something less than 23 maximum out of the possible 255 if I did that right. Second, you'd be more accurate to do the division after you've fully accumulated the 9 weighted pixel values. The way you have it, you get a little truncation for each of the 9. If you wait until the end, you only get the truncation once. So get rid of the division by 16 in the loop and replace "size" by 16 at the end.
Now to figure out how you're getting the 9 pixels to average. All I know for sure is that I'm 99.9% sure what you're doing isn't right. 
If you don't have the data, you're just another a**hole with an opinion.
|
|
|
|
|
Well, this those changes, it now comes out red and black instead of grey and black.
I'm Sure I'm doing it wrong too.
"Sir, I protest. I am NOT a merry man!"
|
|
|
|
|
If the colors are changing, then what you're doing wrong is your pixel lookup, the colors are bleeding between red, green and blue. Did you read my article which has a guassian blur filter in it ?
Christian Graus
Please read this if you don't understand the answer I've given you
"also I don't think "TranslateOneToTwoBillion OneHundredAndFortySevenMillion FourHundredAndEightyThreeThousand SixHundredAndFortySeven()" is a very good choice for a function name" - SpacixOne ( offering help to someone who really needed it ) ( spaces added for the benefit of people running at < 1280x1024 )
|
|
|
|
|
Yes, I did.
However, I don't see how my pixel lookup is wrong. :S
"Sir, I protest. I am NOT a merry man!"
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
