Format for displaying the number of objects in a zone

[epigraphe]

Search before asking

• I have searched the Supervision issues and found no similar feature requests.

Question

Hello. I re-read the documentation and did not find how to fix this.
Counting objects in a zone looks like the number point zero for me. (1.0 for example)
How to remove .0?

In all your examples, the objects have an integer number, without dots and numbers after the dot

import cv2
import streamlit as st
import numpy as np
from ultralytics import YOLO
import supervision as sv
import time


ZONE_POLYGON = np.array([
    [0, 0],
    [1, 0],
    [1, 1],
    [0, 1]
])

def main():
    st.title("YOLOv8 Live Demo")

#    frame_width, frame_height = st.slider("Select Webcam Resolution (Width, Height)", 0, 1920, 1280, 1), st.slider("", 0, 1080, 720, 1)
    frame_width, frame_height = 1080, 720


    id_cap = cv2.VideoCapture(0)
    id_cap.set(cv2.CAP_PROP_FRAME_WIDTH, frame_width)
    id_cap.set(cv2.CAP_PROP_FRAME_HEIGHT, frame_height)

    model = YOLO("yolov8n.pt")
    model.predict(source="0", show=False, stream=True, classes=0)
    box_annotator = sv.BoxAnnotator(
        thickness=4,
        text_thickness=4,
        text_scale=2
    )

    zone_polygon = (ZONE_POLYGON * np.array([frame_width, frame_height])).astype(int)
    zone = sv.PolygonZone(polygon=zone_polygon, frame_resolution_wh=(frame_width, frame_height))
    zone_annotator = sv.PolygonZoneAnnotator(
        zone=zone, 
        color=sv.Color.white(),
        thickness=6,
        text_thickness=6,
        text_scale=4
    )

    show_video = st.checkbox("Show Video", value=True)

    st_frame = st.empty()

    start_time_hourly = time.time()
    start_time_10_sec = time.time()

    while show_video:
        success, frame = id_cap.read()

        if success:
            result = model(frame, agnostic_nms=True)[0]
            detections = sv.Detections.from_ultralytics(result)

            labels = [
                f"{model.model.names[class_id]} {confidence:0.2f}"
                for _, _, confidence, class_id, _ in detections
            ]

            frame = box_annotator.annotate(
                scene=frame, 
                detections=detections, 
#                labels=labels
                skip_label=True
            )

            zone.trigger(detections=detections)
            zone_value = zone.current_count
            current_hour = int(time.strftime("%H"))
            filename_hourly = f"{current_hour}-{(current_hour + 1) % 24:02}_{time.strftime('%d-%m-%Y')}.txt"
            current_time = time.strftime("%H:%M:%S")
            current_time_10_sec = time.time()
            elapsed_time_10_sec = current_time_10_sec - start_time_10_sec

            if elapsed_time_10_sec >= 10:
                with open(filename_hourly, "a") as file:
                    file.write(f"{current_time} {zone_value}\n")
                start_time_10_sec = time.time() 
            current_time_hourly = time.time()
            elapsed_time_hourly = current_time_hourly - start_time_hourly

            if elapsed_time_hourly >= 3600:
                start_time_hourly = time.time()

            frame = zone_annotator.annotate(scene=frame)

            st_frame.image(frame, channels="BGR")
            
        else:
            id_cap.release()
            break

        


if __name__ == "__main__":
    main()

Additional

No response

[SkalskiP]

Hi, @epigraphe 👋🏻! We fixed that in the current develop. Try to install it like this: pip install supervision==0.15.0rc2, and let us know if the issue persists.

[epigraphe]

Yes fine. After updating to the version you specified, everything works as it should. Thank you very much

[SkalskiP]

Excellent! supervision-0.15.0 should be officially out this week. Then simple pip install supervision will be enough. I'm closing the issue for now.