Update RVC4 FW and add an example

cmake/Depthai/DepthaiDeviceRVC4Config.cmake

@@ -4,4 +4,4 @@ set(DEPTHAI_DEVICE_RVC4_MATURITY "snapshot")
 
 # "version if applicable"
 # set(DEPTHAI_DEVICE_RVC4_VERSION "0.0.1+93f7b75a885aa32f44c5e9f53b74470c49d2b1af")
-set(DEPTHAI_DEVICE_RVC4_VERSION "0.0.1+e15e34e96495e1ac3d4c064499d45f33784cba8a")
+set(DEPTHAI_DEVICE_RVC4_VERSION "0.0.1+0c4703535d47af4edaa3435abd25d4c93b4938d6")

examples/python/Camera/camera_raw.py

@@ -0,0 +1,70 @@
+#!/usr/bin/env python3
+
+import cv2
+import depthai as dai
+import numpy as np
+
+def unpack_raw10(raw_data, width, height):
+    """
+    Unpacks RAW10 data from DepthAI pipeline into a 16-bit grayscale array.
+    :param raw_data: List of raw bytes from DepthAI (1D numpy array)
+    :param width: Image width
+    :param height: Image height
+    :return: Unpacked 16-bit grayscale image
+    """
+    expected_size = (width * height * 10) // 8
+
+    if len(raw_data) < expected_size:
+        raise ValueError(f"Data too small: {len(raw_data)} bytes, expected {expected_size}")
+
+    # Convert raw_data to numpy array
+    packed_data = np.frombuffer(raw_data, dtype=np.uint8)
+
+    # Reshape into groups of 5 bytes (4 pixels per group)
+    packed_data = packed_data[:expected_size].reshape(-1, 5)
+    unpacked_data = np.zeros((packed_data.shape[0], 4), dtype=np.uint16)
+
+    # Extract 8 most significant bits
+    unpacked_data[:, 0] = packed_data[:, 0].astype(np.uint16) << 2
+    unpacked_data[:, 1] = packed_data[:, 1].astype(np.uint16) << 2
+    unpacked_data[:, 2] = packed_data[:, 2].astype(np.uint16) << 2
+    unpacked_data[:, 3] = packed_data[:, 3].astype(np.uint16) << 2
+
+    # Extract least significant 2 bits from 5th byte
+    unpacked_data[:, 0] |= (packed_data[:, 4] & 0b00000011)
+    unpacked_data[:, 1] |= (packed_data[:, 4] & 0b00001100) >> 2
+    unpacked_data[:, 2] |= (packed_data[:, 4] & 0b00110000) >> 4
+    unpacked_data[:, 3] |= (packed_data[:, 4] & 0b11000000) >> 6
+
+    # Reshape to image dimensions
+    raw_image = unpacked_data.flatten().reshape(height, width)
+
+    # Scale from 10-bit (0-1023) to 16-bit (0-65535) for proper display
+    raw_image_16bit = (raw_image * 64).astype(np.uint16)
+
+    return raw_image_16bit
+
+# Create pipeline
+with dai.Pipeline() as pipeline:
+    # Define source and output
+    cam = pipeline.create(dai.node.Camera).build(dai.CameraBoardSocket.CAM_A)
+    videoQueue = cam.raw.createOutputQueue()
+
+    # Connect to device and start pipeline
+    pipeline.start()
+    while pipeline.isRunning():
+        # videoIn = videoQueue2.tryGet()
+        videoIn = None
+        raw = videoQueue.tryGet()
+        if raw is not None:
+            assert isinstance(raw, dai.ImgFrame)
+            data_raw = raw.getData()
+            print(raw.getStride())
+            raw_image = unpack_raw10(data_raw, raw.getStride(), raw.getHeight())
+            cv2.imshow("raw", raw_image)
+        if videoIn is not None:
+            assert isinstance(videoIn, dai.ImgFrame)
+            cv2.imshow("video", videoIn.getCvFrame())
+
+        if cv2.waitKey(1) == ord("q"):
+            break