Intro to Computer Vision Using AForge.NET AForge.NET is a ... › 1633735 ›...

Intro to Computer Vision Using AForge.NET

AForge.NET is a computer vision and artificial intelligence library originally developed by Andrew Kirillov for the .NET Framework.

The source code and binaries of the project are available under the terms of the Lesser GPL and the GPL (GNU General Public License).

We are going to walk through a demo on how to setup a project using Visual Studio. I will be using Visual Studio Express C# 2010 for this tutorial. However any version should work just fine.

To get started go to http://www.aforgenet.com/framework/downloads.html and download the installer.

You will also need to Visual Studio go to https://www.visualstudio.com/en-us/products/visual-studio-express-vs.aspx

Or if you want you can still download Visual Express 2010 here

http://download.microsoft.com/download/1/E/5/1E5F1C0A-0D5B-426A-A603-1798B951DDAE/VS2010Express1.iso

1

https://www.visualstudio.com/en-us/products/visual-studio-express-vs.aspxhttps://www.visualstudio.com/en-us/products/visual-studio-express-vs.aspxhttp://download.microsoft.com/download/1/E/5/1E5F1C0A-0D5B-426A-A603-1798B951DDAE/VS2010Express1.isohttp://download.microsoft.com/download/1/E/5/1E5F1C0A-0D5B-426A-A603-1798B951DDAE/VS2010Express1.iso


Run the Aforgeinstaller. And notethe location of theinstalled framework.

Complete theinstallation and installVisual Studio.

2


Open Visual Studio Express.

Andcreate anew project.

Select Windows Forms Application.

Enter in the Project Name and click OK.

3


First we need to add a reference to the Aforge.NET.

Right click on References and select Add Reference.

Then navigate to the Aforge.NET release folder: C:\Program Files\AForge.NET\Framework\Release

We want to add reference to four libraries.

C:\Program Files\AForge.NET\Framework\Release\AForge.dllC:\Program Files\AForge.NET\Framework\Release\AForge.Imaging.dllC:\Program Files\AForge.NET\Framework\Release\AForge.Video.dllC:\Program Files\AForge.NET\Framework\Release\AForge.Video.DirectShow.dll

4


5


Next double click on Form1.cs in the Solution Explorer menu to the right. Then press F7 to go into view code.

And add the reference to the top of the code.

using AForge;using AForge.Imaging;using AForge.Video;using AForge.Video.DirectShow;

Press Shift-F7 to switch back to view design.

Add three controls to your userform, PictureBox, Button and ComboBox.

Rename each control.

ComboBox –> cboCameraButton → btnStartCameraPictureBox → picVideo

Change the text on the button to“Start Camera” and Set the Sizefor PictureBox to 320,240

Set PictureBox SizeMode toStretchImage

Next double click anywhere onthe form where there is not acontrol. This will create a LoadEvent for the form.

6


Add the following code

// Add Global References private FilterInfoCollection colCamera; private VideoCaptureDevice CaptureDevice;

private void Form1_Load(object sender, EventArgs e) { FillInDropDownCamera(); }

void FillInDropDownCamera() { colCamera = new FilterInfoCollection(FilterCategory.VideoInputDevice); foreach (FilterInfo Device in colCamera) { cboCamera.Items.Add(Device.Name); cboCamera.SelectedIndex = 0; CaptureDevice = new VideoCaptureDevice(); } }

This code will get a list of the cameras available and add them to the drop down ComboBox list.

Press F5 to compile and run

You should see the list of camerasavailable.

Close the userform or press the Stopbutton in Visual Studio

7


Next switch back to the Design Mode and double click on the Start Camera Button.

This will make an event for the button.

We want to add the code to the button

private void btnStartCamera_Click(object sender, EventArgs e) { CaptureDevice = new VideoCaptureDevice(colCamera[cboCamera.SelectedIndex].MonikerString); CaptureDevice.NewFrame += new NewFrameEventHandler(FinalFrame_NewFrame); CaptureDevice.Start(); }

void FinalFrame_NewFrame(object sender, NewFrameEventArgs eventArgs) {

Bitmap video1 = (Bitmap)eventArgs.Frame.Clone(); picVideo.Image = video1;

}

The click event will start the camera and the NewFrameEventHandler will run everytime a frame of video is captured. The video is then overlayed on the PictureBox.

Next we need to add two more events, We will start with a FormClosed event.

Select the border of the userform and click on the lightning bolt icon at the top of properties menu. Find the FormClosed event and double click next to it to add the event,

8


Next add an event for the ComboBox DropDown.

Select the ComboBox and select the lightning bolt. Double Click on SelectedIndexChanged

You will now see the two events added to your code.

Add the following code. This will ensure that the camera closes on exit. And when the ComboBox changes selection.

private void Form1_FormClosed(object sender, FormClosedEventArgs e) { CloseCamera(); }

private void cboCamera_SelectedIndexChanged(object sender, EventArgs e) { CloseCamera(); }

void CloseCamera() {

if (CaptureDevice != null) { if (CaptureDevice.IsRunning == true) { CaptureDevice.Stop(); picVideo.Image = new Bitmap(640, 480); } } }

9


Press F5 to run your code.

Next let's add a second PictureBox to our Userform.

You can select the picturebox and copy and paste a copy with CTRL C and CTRL V.

Move the box to the right and rename to picVideoProcessed.

Add a line to the CloseCamera subroutine to blank out the video on close

picVideoProcessed.Image = new Bitmap(640, 480);


if (CaptureDevice != null) { if (CaptureDevice.IsRunning == true) { CaptureDevice.Stop(); picVideo.Image = new Bitmap(640, 480); → picVideoProcessed.Image = new Bitmap(640, 480); } } }

10


We need to add a reference to the second PictureBox to display the image.

Add the following code

Bitmap video2 = (Bitmap)eventArgs.Frame.Clone();picVideoProcessed.Image = video2;

picVideoProcessed.Image = new Bitmap(640, 480);


Bitmap video1 = (Bitmap)eventArgs.Frame.Clone(); picVideo.Image = video1; → Bitmap video2 = (Bitmap)eventArgs.Frame.Clone(); → picVideoProcessed.Image = video2;

}


if (CaptureDevice != null) { if (CaptureDevice.IsRunning == true) { CaptureDevice.Stop(); picVideo.Image = new Bitmap(640, 480); → picVideoProcessed.Image = new Bitmap(640, 480); } } }

11


Now we can apply different filters to affect the processed video.

Each frame is process separately as the FinalFrame_NewFrame event is triggered.

Add a new filter

System.Drawing.Image MirrorImage(System.Drawing.Image Image1) { // create filter Mirror filter = new Mirror(false, true); // apply the filter filter.ApplyInPlace((Bitmap)Image1); return Image1; }

And edit the line

picVideoProcessed.Image = video2;

to

picVideoProcessed.Image = MirrorImage(video2);

12


Finally Let's Add a checkbox to our form and name it chkRaw and add the following code.

This code will try to track an object based on it's color. We can run it inline with other filters.

int intRedLower = 50; int intRedUpper = 100; int intGreenLower = 0; int intGreenUpper = 50; int intBlueLower = 0; int intBlueUpper = 50;

System.Drawing.Image CameraTracking(Bitmap video2) {

// create filter ColorFiltering colorFilter = new ColorFiltering(); // set color ranges to keep colorFilter.Red = new IntRange(intRedLower, intRedUpper); colorFilter.Green = new IntRange(intGreenLower, intGreenUpper); colorFilter.Blue = new IntRange(intBlueLower, intBlueUpper);

// apply the filter colorFilter.ApplyInPlace(video2); BlobCounter BlobCounter = new BlobCounter();

// MinHeight doesn't seem to do anything? BlobCounter.MinHeight = 50;

BlobCounter.ObjectsOrder = ObjectsOrder.Size;

BlobCounter.ProcessImage(video2); Rectangle[] rect = BlobCounter.GetObjectsRectangles(); if (chkRaw.Checked) { if (rect.Length > 0) { for (int i = 0; i


{ bool bolRectFound = false;

if (rect.Length > 0) { Rectangle obj = rect[0]; int X = obj.X + (obj.Width / 2); int Y = obj.Y + (obj.Height / 2);

// filter out small objects int intThreshhold = 5; if (obj.Width > intThreshhold && obj.Height > intThreshhold) {

Graphics graphic = Graphics.FromImage(video2); using (Pen Pen = new Pen(Color.White, 3)) { graphic.DrawRectangle(Pen, obj); } graphic.Dispose(); bolRectFound = true; }

}

}

return video2; }

and then we can run the code on each frame.


Bitmap video1 = (Bitmap)eventArgs.Frame.Clone(); picVideo.Image = video1; Bitmap video2 = (Bitmap)eventArgs.Frame.Clone(); → picVideoProcessed.Image = CameraTracking((Bitmap) MirrorImage(video2));

}

You can adjust the color ranges by changing the intRefLower and intRedUpper Ranges.

14

Date post:	27-Jan-2021
Category:	Documents
Upload:	others
View:	5 times
Download:	0 times

Intro to Computer Vision Using AForge.NET AForge.NET is a ... › 1633735 ›...

Documents