Skip to content

Latest commit

 

History

History

Python NET

README.md

Python samples for Windows NET

Requirements

Maybe newer versions of Python are supported by PythonNET now.

These samples show how to use IC Imaging Control 3.5.

The DLLs are located in the dist directory of "Documents/IC Imaging Control 3.5". This is saved in the environment varialbe "IC35PATH". For PythonNET being able to find the DLLS, we simply add this path to the search path:

import sys, os
sys.path.append(os.getenv('IC35PATH') + "/redist/dotnet/x64")

The qt5-simple.py sample simply displays the live video in a QT5 widget. For doing so, a frame filter is inserted into the display path of IC Imaging Control. The filter converts the IFrame of IC Imaging control into a QImage and mirrors it. Then a signal is sent to the main thread.

The qt5-imageprocessing.py sample enhances the qt5-simple.py with a frame callback and some simple image processing. The results are signaled to the main thread again.

The snap-save-image.py sample shows in a very simply way how to

  • open a video capture device
  • save and restore a device state of a video capture device
  • set automatic and value properties, such as gain and exposure
  • snap a frame from the live stream
  • save the snapped frame into a JPEG file.

The sample has no GUI, therefore, there is no live display from the video capture device shown.

Camera Properties

The usual VCD Properties of IC Imaging Control are used here. Please refer to VCDProperties

If IC Imaging Control .NET SDK is installed on the computer, then the VCDProperty Inspector program can be used to query the VCDIDs, VCDElements and VCDInterfaces of the camera properties.

Getting an interface to a property is simple:

# Get the Gain value.
GainValue = ic.VCDPropertyItems.FindInterface(TIS.Imaging.VCDIDs.VCDID_Gain,
                TIS.Imaging.VCDIDs.VCDElement_Value,TIS.Imaging.VCDIDs.VCDInterface_Range)

Setting the value is also simple:

# Set minimum gain
if GainValue is not None:
    GainValue.Value = 10

Setting the minumum value or maximim value:

# Set minimum gain
if GainValue is not None:
    GainValue.Value = GainValue.RangeMin

This is similar for all numeric video capture device properties. Take care of properties, that also have an automatic. That must be disabled before a value can be set manually. For gain this is:

# Get the Gain automatic property
GainAuto = ic.VCDPropertyItems.FindInterface(TIS.Imaging.VCDIDs.VCDID_Gain,
                TIS.Imaging.VCDIDs.VCDElement_Auto,TIS.Imaging.VCDIDs.VCDInterface_Switch)

# Disable Gain automatic                                
if GainAuto is not None:
    GainAuto.Switch = False

Snap and save a frame

For snapping a frame we use the FrameSnapSink. It is created and passed to the IC Imaging Control instance before the live video stream is started. This can be done even before a video capture device is opened.

snapsink = TIS.Imaging.FrameSnapSink(TIS.Imaging.MediaSubtypes.RGB32)

The pixel format in the sink is RGB32 as we passed this format to the constructor. Now the sink must be assigned to the IC Imaging Control instance:

ic.Sink = snapsink

Now the video capture device can opened, if not already done. The live stream can be started now.

A frame is snapped from the live stream by a call to:

frame = snapsink.SnapSingle(TimeSpan.FromSeconds(5))

This frame can be converted to a numpy array (see the QT5 samples) or we can simply save it to a file:

# Save the frame as JPG with 75% quality.
TIS.Imaging.FrameExtensions.SaveAsJpeg( frame,"image.jpg", 75)

Capture a Video File

Sample: qt5-mediastreamsink

Copied from qt5-simple.py for enhancement.

This sample is based on the C# "Capture a video file", which is documented at [Capturing a Video File]https://www.theimagingsource.com/support/documentation/ic-imaging-control-net/CaptureVideo.htmFrameSnapSink.htm). It is quite simple, because as soon as the live stream is started, the video is capture to a file "test.mp4".

The sample shows, how to enumerate the stream containers and select the MP4 container. It also shows, how to use a lambda function in order to select the MediaFoundation h.264" codec, which is supported by the MP4 StreamContainer. At start, the property dialog of the codec is shown.

Without error handling, the creation of the MediaStreamSink looks as follows:

# Get MP4 stream container
CurrentMediaStreamContainer = None
for container in TIS.Imaging.MediaStreamContainer.MediaStreamContainers:
    if container.Name == "MP4":
        CurrentMediaStreamContainer = container

# Get the MediaFoundation h.264 codec
try:
    CurrentCodec = next(c for c in TIS.Imaging.AviCompressor.AviCompressors 
                        if CurrentMediaStreamContainer.IsCodecSupported(c) and 
                        c.Name == "MediaFoundation h.264")
except:
    print("Codec not found!)
    quit()    

Filename = "test." + CurrentMediaStreamContainer.PreferredFileExtension
MediaStreamSink = TIS.Imaging.MediaStreamSink(CurrentMediaStreamContainer, CurrentCodec[0],
                                              Filename)

ic.Sink = MediaStreamSink

However, always add error handling, as shown in the Python file! Also: ic.Sink has a capital "S" at "Sink". I write this, because I used "ic.sink" and it took me half an hour to find that error.

In order to keep the sample as simple as possible, the user interface does not contain file selection, start and stop functions. For enhancing the sample, please refer to the above mentioned link.