Merge pull request #5 from SICKAG/grab_numpy_example

Grab numpy example

Author: olofsjo1SICKAG

README.md

@@ -1,10 +1,14 @@
 ## About
 The purpose of this repository is to give a starting point and some code examples in Python on what can be done using [Harvesters](https://github.com/genicam/harvesters/) during integration with a [Ranger/Ruler device](https://www.sick.com/se/en/catalog/products/machine-vision-and-identification/machine-vision/ranger3/c/g448354).
 
-Since our GenIStream API does not have Python support, we recommend Harvesters using for integration using Python and a Ranger/Ruler device. 
+Since our GenIStream API does not have Python support, we recommend Harvesters using for integration using Python and a Ranger/Ruler device.
 
 ### Version
 Currently tested with Harvesters 1.4.2 and Python 3.11.1.
 
 ### CTI-file
 Get the CTI-file from the SDK, it is located at: `"GenTL producer/x64-release/SICKGigEVisionTL.cti"`. This file is the one you want to load into Harvesters.
+
+## Import an existing icon-image (dat-file)
+If you instead want to use existing images saved to disk you can have a look at this article at the support portal (need login):
+https://support.sick.com/sick-knowledgebase/article/?code=KA-07788

ranger3_grab_process.py

@@ -0,0 +1,123 @@
+from harvesters.core import Harvester
+import numpy as np
+import matplotlib.pyplot as plt
+import os
+
+
+def reshape_component(component, type=np.uint16):
+    h, w = component.height, component.width
+    image = component.data.reshape(h, w)
+
+    return np.array(image, dtype=type)
+
+
+def get_components(buffer):
+    """
+    Extracts the components from the buffer and reshapes them into numpy arrays.
+    :param buffer: The buffer containing the components.
+    :return: A dictionary containing the reshaped components.
+    """
+
+    # The following keys can be used to identify a component:
+    # 3300: Region ID
+    # 3301: Component ID
+
+    # Region ID:
+    # 11: Extraction region 1
+    # 16: Dual Exposure Extraction region
+    # 12: Extraction region 2
+    # 13: Extraction region 3
+    # 14: Extraction region 4
+    # 15: Extraction region 5
+    # 17: Extraction White
+    # 18: Extraction Color
+
+    # Component ID:
+    # 1: Intensity (2D)
+    # 4: Range
+    # 5: Reflectance
+    # 7: Scatter
+    components = {}
+    for component in buffer.payload.components:
+        region_id = component._part.get_info_int64(3300)
+        component_id = component._part.get_info_int64(3301)
+
+        if region_id == 11:  # Extraction region 1
+            if components.get("Scan3dExtraction1") is None:
+                components["Scan3dExtraction1"] = {}
+            if component_id == 4:  # Ranger
+                components["Scan3dExtraction1"]["Range"] = reshape_component(component, np.uint16)
+            elif component_id == 5:  # Reflectance
+                components["Scan3dExtraction1"]["Reflectance"] = reshape_component(component)
+            elif component_id == 7:  # Scatter
+                components["Scan3dExtraction1"]["Scatter"] = reshape_component(component)
+            else:
+                print("Unknown component ID: ", component_id)
+        else:
+            print("Unknown region ID: ", region_id)
+
+    return components
+
+
+def main():
+    # Create the Harvester instance
+    h = Harvester()
+
+    # Load the cti-file
+    cti_path = os.path.join(os.path.dirname(__file__), "SICKGigEVisionTL.cti")
+    h.add_file(cti_path, check_existence=True, check_validity=True)
+
+    # Update Harvester, this will update the list of devices
+    h.update()
+    print("Found devices: ", [(device.user_defined_name, device.serial_number) for device in h.device_info_list])
+
+    # The following keys can be used to identify a device:
+    # - list_index: item index of the list of :class:`DeviceInfo` objects.
+    # - id_: Index of the device information list.
+    # - vendor: Vendor name of the target device.
+    # - model: Model name of the target device.
+    # - tl_type: Transport layer type of the target device.
+    # - user_defined_name: User defined name string of the target device.
+    # - serial_number: Serial number string of the target device.
+    # - version: Version number string of the target device.
+    # Note: If list_index is specified that is always used and the other keys
+    # are not used, otherwise a combination of the others shall be used to
+    # specify an unique device
+    with h.create({"user_defined_name": "ZPM03"}) as ia:
+
+        node_map = ia.remote_device.node_map
+
+        # Set to 3d mode
+        node_map.DeviceScanType.value = 'Linescan3D'
+
+        # Start the acquisition
+        ia.start()
+
+        # Fetch data from the image
+        print("===== Waiting for image =====")
+        with ia.fetch() as buffer:
+            print("===== Image Aquired =====")
+            components = get_components(buffer)
+            region1_range = components["Scan3dExtraction1"]["Range"]
+            region1_reflectance = components["Scan3dExtraction1"]["Reflectance"]
+            region1_scatter = components["Scan3dExtraction1"]["Scatter"]
+
+        ia.stop()
+
+        print("===== Camera Stopped =====")
+
+        # Save the data to numpy files
+        np.save("data/region1_range", region1_range)
+        np.save("data/region1_reflectance", region1_reflectance)
+        np.save("data/region1_scatter", region1_scatter)
+
+        print(region1_reflectance)
+        plt.subplot(2, 1, 1)
+        plt.imshow(region1_reflectance, interpolation='none')
+        plt.subplot(2, 1, 2)
+        plt.imshow(region1_range, interpolation='none')
+        plt.show()
+
+
+if __name__ == "__main__":
+    main()