Added vision_stack package

SubjuGator/perception/subjugator_perception/nodes/vision_stack_test.py

@@ -0,0 +1,38 @@
+#!/usr/bin/env python3
+import rospy
+from image_geometry import PinholeCameraModel
+from mil_ros_tools import (
+    Image_Subscriber,
+)
+from vision_stack import ResizeLayer, UnderWaterImageEnhancementLayer, VisionStack
+
+__author__ = "Daniel Parra"
+
+
+class ObjectDetectionTest:
+    def __init__(self):
+        camera = rospy.get_param("~image_topic", "/camera/front/right/image_rect_color")
+        SIZE = (960, 608)
+        self.vs = VisionStack(
+            layers=[
+                ResizeLayer((0, 0), 960, 608),
+                UnderWaterImageEnhancementLayer(SIZE),
+            ],
+            input_size=SIZE,
+        )
+
+        self.image_sub = Image_Subscriber(camera, self.detection_callback)
+        self.camera_info = self.image_sub.wait_for_camera_info()
+        assert self.camera_info is not None
+        self.cam = PinholeCameraModel()
+        self.cam.fromCameraInfo(self.camera_info)
+
+    def detection_callback(self, msg):
+        # Create Image from array
+        self.vs.run(msg, True)
+
+
+if __name__ == "__main__":
+    rospy.init_node("vision_pipeline_test")
+    ObjectDetectionTest()
+    rospy.spin()

ruff.toml

@@ -20,6 +20,7 @@ extend-exclude = [
     "NaviGator/simulation/VRX/vrx",
     "NaviGator/simulation/VRX/vrx-docker",
     "mil_common/perception/yolov7-ros",
+    "mil_common/perception/vision_stack",
     # sub simulation
     "SubjuGator/simulation",
     # docker contents

scripts/setup.bash

@@ -114,4 +114,4 @@ cmonly() {
 alias xbox=startxbox
 
 # PYTHONPATH modifications
-export PYTHONPATH="${HOME}/catkin_ws/src/mil/mil_common/axros/axros/src:${PYTHONPATH}"
+export PYTHONPATH="${HOME}/catkin_ws/src/mil/mil_common/perception/vision_stack/src/vision_stack:${HOME}/catkin_ws/src/mil/mil_common/axros/axros/src:${PYTHONPATH}"