|
please 
|
|
|
|
|
|
Hi every one
I need a help in JPEG image processing with C# language ,I need to know how can I read the jpeg file bit’s in the stream wile it compressed and get from it each pixel value in jpeg pixel format and then convert this pixel from jpeg format into RGB format and return it for the jpeg format ,if that is possible.
I really need a help on this I search a lot for this but nothing I get ,if any one have any idea or resource about this problem pleas tell me about it ,if you can.
memo
|
|
|
|
|
What you're asking is kind of bizarre. A JPEG is compressed, you can't read it bit by bit to get the pixel values, you need to decompress it. Luckily, the Bitmap::FromFile function will do this for you for free. All you need to do is load the jpeg from disc, and then use the approach seen in my image processing articles here on code project to access the bits. Remember, if you're going to edit a jpg, jpg has lossy compression, so some detail is lost when you save it again.
To summarise, you need to load the entire jpeg, edit it as an in memory bitmap, and save it again as a jpeg. No compression scheme will allow you to change one pixel without changing significantly more than one byte of the image in it's compressed format.
Christian
I have several lifelong friends that are New Yorkers but I have always gravitated toward the weirdo's. - Richard Stringer
|
|
|
|
|
hi
thanks you very much for answering me.
I will explain my problem , and why I want to do this bizarre  thing ,this problem come from my project where I try to hide some kind of data in all the byte’s in the image pixel’s without insert any noise for the image colour and it worked with me in the bitmap format (RGB) ,but I try now to drive my algorithm to work on the jpeg format either but I found a lot of problem on this that because.
I couldn’t understand for this time how the jpeg compressed ,and how it store each compressed pixel in the jpeg file ,and how could I get this stored pixel and returned it into RGB format and work my algorithm with it and then return it into JPEG format and store it that only if that could be possible ,you can see that my problem come from ,that I still try to understand the JPEG image format but I tried very much to solve this problem but nothing helped me because I couldn’t find a good resource about this subject.
So as you said “If you're going to edit a jpg, jpg has lossy compression, so some detail is lost when you save it again “ and I want for this to be happen because any bit I loss it could cause for me a disaster.
I don’t ask here to tell me if there is any ready library can help me to do this I just want to know does that possible to be happen or not? And if it can be , how??
I believe that all the programming is just a game nothing more or less.
Thanks again for every one give me a suggestion that can help me to solve this bizarre problem.
i try to learn programming what ever i had i feal that i did'nt get anything
|
|
|
|
|
|
memoboy wrote:
..I try to hide some kind of data..
You mean steganography?
Corinna John has written a long series of articles on that subject here on CP: Clicky[^]
--
Help me! I'm turning into a grapefruit!
Phoenix Paint - back from DPaint's ashes!
|
|
|
|
|
you what are you saying about the steganography and what corinna john did about it she provide a very wonderful work ,I have been read all of it work and I like it very much but in my project I use another algorithm which allow for me to hide a huge mountain of bit's in each pixel without insert any noise on the image it self ,but as I have seen in corinna work the passward control the a mount of bit's that you can hide in the image ,my algorithm allow for me to hide in all the pixel.
And that is a very big different here.
i try to learn programming what ever i had i feal that i did'nt get anything
|
|
|
|
|
I guessed this is what you were doing, image masking is why I almost never use jpeg, for exactly this reason.
Christian
I have several lifelong friends that are New Yorkers but I have always gravitated toward the weirdo's. - Richard Stringer
|
|
|
|
|
And to add what Christian said, use Bitmap.LockBits to get the BitmapData in which you process each pixel, and Bitmap.UnlockBits to release the lock (locks the data in memory so it isn't moved while processing it) on the bits.
This posting is provided "AS IS" with no warranties, and confers no rights.
Software Design Engineer
Developer Division Sustained Engineering
Microsoft
[My Articles] [My Blog]
|
|
|
|
|
*grin* I was hoping he'd get that bit from my articles
Christian
I have several lifelong friends that are New Yorkers but I have always gravitated toward the weirdo's. - Richard Stringer
|
|
|
|
|
ok but as i know if i open the image and modefy it bit's in bitmabe formate and then i restore it as jpeg foramte there will be a lose in bit's value
my project need from me to keep some spacifice bit's in each pixel with keeping it in it origonal form without losing it there is any way for that
memo
|
|
|
|
|
memoboy wrote:
ok but as i know if i open the image and modefy it bit's in bitmabe formate and then i restore it as jpeg foramte there will be a lose in bit's value
Yep, like I said, that's what 'lossy compression' means.
memoboy wrote:
my project need from me to keep some spacifice bit's in each pixel with keeping it in it origonal form without losing it there is any way for that
You want jpeg compression to be lossless ? No, you can't achieve that, you need to use a different image format.
Christian
I have several lifelong friends that are New Yorkers but I have always gravitated toward the weirdo's. - Richard Stringer
|
|
|
|
|
JPEG2000 claims to support lossless compression, but GDI+ does not support it yet in any release bits. If one could find a GDI+ image codec then .NET would inherit the format by instantiating your own ImageFormat using whatever Guid defined with such a codec.
To the original poster, if you want lossless compression PNG is about as close as you'll get.
You can also control the amount of compression (less compression == better quality == larger file size) by using the Save override to specify an EncoderParameters object and using the EncoderValue enumeration with one of the compression members if you save as a TIFF image (.tif or .tiff).
This posting is provided "AS IS" with no warranties, and confers no rights.
Software Design Engineer
Developer Division Sustained Engineering
Microsoft
[My Articles] [My Blog]
|
|
|
|
|
hi every 1 ,
i'm working on a simple networking project and i want to transfer files between the client and the server , i use a FileStram and StreamReader in the client to read the data from that file . but i think i no data are retreved from thet file . is there any way to do that correctly ?! can any 1 write a few lines to show me how to get this byts and send them thrugh the network ?
please help me !
|
|
|
|
|
You need to use a NetworkStream, which is documented in the .NET Framework SDK class library[^].
You can obtain a NetworkStream from using a TcpClient or a plain Socket connection. You can also obtain one from WebResponse.GetResponseStream. Pluggable protocol handlers that extend WebRequest and WebResponse like the HttpWebRequest and HttpWebResponse class, respectively, communicate across channels (like HTTP for the HttpWebRequest class, obviously) and allow you to stream both request and response data.
Look at the classes available in the System.Net and System.Net.Sockets namespaces in the .NET Framework, which I've linked above and is installed by default with VS.NET. It's also available as a separate download from http://msdn.microsoft.com/netframework[^].
If all you need to do is copy a file from a network share you can use File.Copy. If the user has permissions the file will copy across the network to the destination location just like how files are copied when using Windows. If the user does not have permissions and you need to impersonate a use who does (which requires that you prompt for credentials), then read about the WindowsIdentity.Impersonate[^] method.
This posting is provided "AS IS" with no warranties, and confers no rights.
Software Design Engineer
Developer Division Sustained Engineering
Microsoft
[My Articles] [My Blog]
|
|
|
|
|
I have a table that I require calculations to be performed against.
To achieve this, I have written several methods for the required functions.
I have no problem with this solution although it does appear limited.
As an example, I have created a form which is passed the DataTable. The form allows the user to choose the required function EG Divide, Mltiple, Add or Subtract and then asks what value/column the function should be performed against, what factor value/column should be used in the calculation and finally, asks where the result should be stored.
Now if the user wants to Divide the value in columnA by the value in columnB and store the result somewhere, it all works fine.
But imagine if the user wants to perform a further calculation on the result, at present the result has to be stored somewhere else and a second pass performed using the desired function.
The only way I have though about solving this problem is to create some kind of math object which can be lnked in a daisy-chain or stored in an array so that multiple calculations can be performed therfore creating an equation.
Has anyone had any experience of doing this or similar?
Any ideas on removing the restrictions of the curent design?
Regards
Wayne Phipps
____________
Time is the greatest teacher... unfortunately, it kills all of its students
LearnVisualStudio.Net
|
|
|
|
|
Using a DataTable is rather inefficient with lots of overhead you don't need for a simple state machine and what could be done using simple state variables. Store your result in a variable just like you do for your controls you dropped onto the form in the designer. Read the operands from the form, calculate, and store the result in the variable. If you want to remember all the results use a growable list like ArrayList or Stack. A stack is what your more advanced calculators use to store previous results, and stacks are what are typically used by calculators to evaluate expressions. You can find lots of information on the web about evaluating expressions to perform calculations.
Even if you want to stick with using a DataTable add a column for metadata about the operands and/or operators you have stored in the DataRows. You could, for example, push the result of a calculation into a new DataRow with information that states whether the data is an operand, operator, or result.
This posting is provided "AS IS" with no warranties, and confers no rights.
Software Design Engineer
Developer Division Sustained Engineering
Microsoft
[My Articles] [My Blog]
|
|
|
|
|
I am currently designing a solution for a school to maintain schedules and other stuff online. I want to use windows security for the users.
How can I ask for the windows login data in a windows forms application (to provides credentials for the server)? This because they use Novell, and the server will be on a separate domain with it's own users.
WM.
What about weapons of mass-construction?
|
|
|
|
|
|
Looks great until now.
My next question is: Does the DPAPI also work on win 98?
WM.
What about weapons of mass-construction?
|
|
|
|
|
If you look-up the native APIs that are P/Invoked in the sample code in the MSDN Library[^] like CryptProtectData[^] you'll see that DPAPI isn't, but LogonUser is. Again, though, make sure that you never store nor transmit the password in clear text. LogonUser can authenticate (with or without returning a security token, which you don't need if all you need to do is authenticate the user and not authorize any secure contexts) and will encrypt the data before transmitting it, so run it locally.
If you must support Windows (as opposed to Windows NT, on which NT4, 2000, XP, 2003, and all newer Windows platforms are based) then you'll find that your application will be very limited in terms of security and cryptographic APIs.
This posting is provided "AS IS" with no warranties, and confers no rights.
Software Design Engineer
Developer Division Sustained Engineering
Microsoft
[My Articles] [My Blog]
|
|
|
|
|
I have to support windows 98 since that's the OS the clients will be running on. It would be better to say: Use windows XP or else you are soooo screwed. But yeah, customers ... 
Anyway, the solution works perfectly
So thanks for the help.
WM.
What about weapons of mass-construction?
|
|
|
|
|
I am looking to allow the user to move a control on the screen by clicking and dragging a control and then releasing the mouse. Much like a desktop shortcut. My control will be a usercontrol and I am unsure of how to understand how to interpret the mouse poisition on screen and on the control along with the mouse event listeners to successfully move the control fluidly around the screen.
Thanks
M4tt
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
