object_detection.py

"""TODO: Add docstring."""

import numpy as np
import pyarrow as pa
from adora import AdoraStatus
from ultralytics import YOLO

CAMERA_WIDTH = 640
CAMERA_HEIGHT = 480


model = YOLO("yolov8n.pt")


class Operator:
    """Inferring object from images."""

    def on_event(
        self,
        adora_event,
        send_output,
    ) -> AdoraStatus:
        """TODO: Add docstring."""
        if adora_event["type"] == "INPUT":
            frame = (
                adora_event["value"].to_numpy().reshape((CAMERA_HEIGHT, CAMERA_WIDTH, 3))
            )
            frame = frame[:, :, ::-1]  # OpenCV image (BGR to RGB)
            results = model(frame, verbose=False)  # includes NMS
            # Process results
            boxes = np.array(results[0].boxes.xyxy.cpu())
            conf = np.array(results[0].boxes.conf.cpu())
            label = np.array(results[0].boxes.cls.cpu())
            # concatenate them together
            arrays = np.concatenate((boxes, conf[:, None], label[:, None]), axis=1)

            send_output("bbox", pa.array(arrays.ravel()), adora_event["metadata"])

        return AdoraStatus.CONTINUE