Hello! When testing the rgb_preview.py script in the examples folder, I noticed the OAK-D-SR outputs a slightly pink tinted image stream compared to the OAK-D-S2. Is this expected? I checked that both are outputting RGB images, so they both seem to be in the same color schema. Should I expect roughly the same colored image between the OAK-D-SR and OAK-D-S2, or am I doing something wrong in the code when it comes to specifying the input nodes?

Here's my minimal code for the OAK-D-SR camera below. Thank you!

import cv2
import depthai as dai

# Create pipeline
pipeline = dai.Pipeline()

# Define source and output
camRgb = pipeline.create(dai.node.ColorCamera)
xoutRgb = pipeline.create(dai.node.XLinkOut)

camRgb.setBoardSocket(dai.CameraBoardSocket.CAM_B)

# Properties
camRgb.setPreviewSize(300, 300)
camRgb.setInterleaved(False)
camRgb.setColorOrder(dai.ColorCameraProperties.ColorOrder.RGB)

xoutRgb.setStreamName("rgb")

# Linking
camRgb.preview.link(xoutRgb.input)

# Connect to device and start pipeline
with dai.Device(pipeline) as device:

    print('Connected cameras:', device.getConnectedCameraFeatures())
    # Print out usb speed
    print('Usb speed:', device.getUsbSpeed().name)
    # Bootloader version
    if device.getBootloaderVersion() is not None:
        print('Bootloader version:', device.getBootloaderVersion())
    # Device name
    #print('Device name:', device.getDeviceName(), ' Product name:', device.getProductName())

    # Output queue will be used to get the rgb frames from the output defined above
    qRgb = device.getOutputQueue(name="rgb", maxSize=4, blocking=False)

    while True:
        inRgb = qRgb.get()  # blocking call, will wait until a new data has arrived

        # Retrieve 'bgr' (opencv format) frame
        cv2.imshow("rgb", inRgb.getCvFrame())

        if cv2.waitKey(1) == ord('q'):
            break

Yes that’s it!

I have a follow up question. For collecting image data for a custom object detection model, does it make more sense then to collect images with the OAK-D-S2 camera or the OAK-D-SR camera for training data purposes? Or will the OAK-D-SR camera “see” differently as well?

    Hi @DiproChakraborty
    Can you send an image of the tinted sensor frames? Does it change with more IR light (eg if you point it towards a window)?

    DiproChakraborty I have a follow up question. For collecting image data for a custom object detection model, does it make more sense then to collect images with the OAK-D-S2 camera or the OAK-D-SR camera for training data purposes? Or will the OAK-D-SR camera “see” differently as well?

    The sensor is different therefore the images will differ slightly, but shouldn't be a problem.

    Thanks,
    Jaka

      Thanks for your help Jaka! Here are some images of the tinted frame from an OAK-D-SR, and I paired it with an image from an OAK-D-S2 for contrast so that you can see the difference. To answer your question, yes, I noticed the OAK-D-SR seems to output more tinted frames when pointed toward a window where the sun is visible.

      So although the images are different, you would suggest that a model (YOLOv8 let's say) trained on images taken by an OAK-D-SR would work on an OAK-D-S2, and vice versa?

      OAK-D-SR

      OAK-D-S2

        Hi @DiproChakraborty
        Looks like a missing IR filter rather than the issue mentioned in the blog post. Please write an email to support@luxonis.com so we can replace it.

        DiproChakraborty So although the images are different, you would suggest that a model (YOLOv8 let's say) trained on images taken by an OAK-D-SR would work on an OAK-D-S2, and vice versa?

        When the sensor issue is solved, sure.

        Thanks,
        Jaka