Does boxmot now support multi target tracking?

[HugTibers]

I saw the author say boxmot doesn't support multiple target last month

[mikel-brostrom]

What do you mean by multiple targets?

[HugTibers]

For example tracking every person waking in the image simultaneously, use from boxmot import DeepOCSORT.

[mikel-brostrom]

This package has always supported multi-object tracking

[HugTibers]

.. ok

[HugTibers]

Hello, its my code

import torch
import cv2
import numpy as np
from boxmot import DeepOCSORT
from pathlib import Path
import os


names = ['fall', 'stand','一次性快餐盒', '充电宝', '垃圾桶', '易拉罐', '瓶子', '筷子']

model = torch.hub.load('./yolov5', 'custom', 'weights/8.pt', source='local')
tracker = DeepOCSORT(
    model_weights=Path('weights/osnet_x0_25_msmt17.pt'),
    device='cuda:0',
    fp16=True,
)

video = 'test/bad.mp4'
cap = cv2.VideoCapture(video)
CountTime = 0

# 追踪
target_id = 1
color = (0, 0, 255)  # BGR
thickness = 2
fontscale = 0.75


# 保存
folder = "runs"
extension = "mp4"
file_number = 1
while True:
    filename = os.path.join(folder, f"{file_number}.{extension}")
    if not os.path.exists(filename):
        break
    file_number += 1

# 基本cv操作
output_path = filename
fourcc = cv2.VideoWriter_fourcc(*'XVID')
fps = int(cap.get(cv2.CAP_PROP_FPS))
frame_size = (int(cap.get(cv2.CAP_PROP_FRAME_WIDTH)), int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT)))
out = cv2.VideoWriter(output_path, fourcc, fps, frame_size)
    
while cap.isOpened():
    ret, frame = cap.read()
    if not ret:
        break
    
    frame_rgb = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
    results = model(frame_rgb)
    CountTime += 1
    
    # 获取长方形和类
    df = results.pandas().xyxy[0]
    print(df)
    if not df.empty:
        xmins = results.pandas().xyxy[0]['xmin']
        ymins = results.pandas().xyxy[0]['ymin']
        xmaxs = results.pandas().xyxy[0]['xmax']
        ymaxs = results.pandas().xyxy[0]['ymax']
        class_list = results.pandas().xyxy[0]['class']
        confidences = results.pandas().xyxy[0]['confidence'] 
        
        newList = []
        for xmin, ymin, xmax, ymax, classitem, conf in zip(xmins, ymins, xmaxs, ymaxs, class_list, confidences):
            xmin = int(xmin)
            ymin = int(ymin)
            xmax = int(xmax)
            ymax = int(ymax)
            
            newList.append([xmin, ymin, xmax, ymax, conf, classitem])
            
            detection = np.array(newList)
            tracks = tracker.update(detection, frame)
            
            print(tracks[:, 5])
        for track in tracks:
            xyxys = track[0:4].astype('int') # float64 to int
            ids = track[4].astype('int') # float64 to int
            confs = track[5]
            clss = track[6].astype('int') # float64 to int
            
            frame = cv2.rectangle(frame, (xyxys[0], xyxys[1]), (xyxys[2], xyxys[3]), color, thickness)
            cv2.putText(
            frame,
            f'id: {ids}, conf: {conf:.2f}, c: {names[clss]}',
            (xyxys[0], xyxys[1]-10),
            cv2.FONT_HERSHEY_SIMPLEX,
            fontscale,
            color,
            thickness
            )
            if CountTime % 10 == 0:
                print(f'识别到人{names[clss]},置信度{conf:.2f}')
                    
        
    out.write(frame)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

print(f'图像识别完成,保存路径为{filename}')
cap.release()
cv2.destroyAllWindows()

i want to know why the result of print(track[:, 5]) is empty when i print(results.pandas().xyxy[0]) more than two bbox.
like:
only one class, print(track[:, 5]) 's result is [0.87]

two class,print(track[:, 5]) is empty

please help me... @mikel-brostrom 🙏🙏🙏