OpenCV目标跟踪之简单质心跟踪

1.质心跟踪算法工作原理

1. 获取到待跟踪目标的边界框
2. 计算质心分配ID
3. 计算新质心与现有对象质心之间的距离（欧几里得距离）
4. 更新现有对象坐标
5. 若出现无关联的新质心，则添加为新对象
6. N个连续帧，旧对象不能与任何现有对象匹配关联则消掉

2.代码

2.1定义一个质心追踪器CentroidTracking类：

from scipy.spatial import distance as dist
from collections import OrderedDict
import numpy as np

class CentroidTracking():
    def __init__(self,maxDisappeared=50):

        # 达到最大连续帧数删除
        self.maxDisappeared = maxDisappeared
        # 分配ＩＤ
        self.nextObjectId = 0
        # 对象字典
        self.objects = OrderedDict()
        self.disappeared = OrderedDict()

    # 增加新对象函数
    def register(self,centroid):
        self.objects[self.nextObjectId] = centroid
        self.disappeared[self.nextObjectId] = 0
        self.nextObjectId += 1

    # 移除对象函数
    def dergister(self, objectID):
        del self.objects[objectID]
        del self.disappeared[objectID]

    # 更新函数
    def update(self,rects):
        if len(rects) == 0:
            for objectID in list(self.disappeared.keys()):
                self.disappeared[objectID] +=1

                if self.disappeared[objectID] > self.maxDisappeared:
                    self.dergister(objectID)
            return  self.objects

        # 存储质心
        inputCentroids = np.zeros((len(rects), 2),dtype="int")

        # 质心坐标
        for (i, (startX, startY, endX, endY)) in enumerate(rects):
            cX = int((startX +endX)/2.0)
            cY = int((startY + endY) /2.0)
            inputCentroids[i] = (cX, cY)

        if len(self.objects) == 0:
            for i in range(0, len(inputCentroids)):
                self.register(inputCentroids[i])

        else:
            # 获取目标id,质心坐标
            objectIDs = list(self.objects.keys())
            objectCentroids = list(self.objects.values())

            # 计算距离排序
            D = dist.cdist(np.array(objectCentroids),inputCentroids)
            rows = D.min(axis=1).argsort()
            cols = D.argmin(axis=1)[rows]

            usedRows = set()
            usedCols = set()

            for (row,col) in zip(rows, cols):
                if row in usedRows or col in usedCols:
                    continue

                objectID = objectIDs[row]
                self.objects[objectID] = inputCentroids[col]
                self.disappeared[objectID] = 0

                usedRows.add(row)
                usedCols.add(col)
            # 未处理的质心
            unusedRows = set(range(0, D.shape[0])).difference(usedRows)
            unusedCols = set(range(0, D.shape[1])).difference(usedCols)

            if D.shape[0] >= D.shape[1]:
                for row in unusedRows:
                    objectID = objectIDs[row]
                    self.disappeared[objectID] += 1

                    if self.disappeared[objectID] > self.maxDisappeared:
                        self.dergister(objectID)

            else:
                for col in unusedCols:
                    self.register(inputCentroids[col])

        return self.objects

2.2 用Caffe模型追踪人脸：

from centroidtracking import CentroidTracking
from imutils.video import VideoStream
from imutils.video import FileVideoStream
import numpy as np
import imutils
import cv2
import time

# 参数路径
prototxt_path = "./deploy.prototxt"
model_path = "./res10_300x300_ssd_iter_140000_fp16.caffemodel"
video_path = "./test.mp4"

centTrack = CentroidTracking()
(H , W) = (None, None)

print("loading model ...")
net = cv2.dnn.readNetFromCaffe(prototxt_path,model_path)

print("starting video stream ...")
# 视频读取
vs = FileVideoStream(video_path).start()
# 摄像头读取
# vs = VideoStream(src=0).start()
time.sleep(1.0)
mean_value =(104, 177, 123)

while True:
    frame =vs.read()
    frame = imutils.resize(frame, width=360)

    if W is None or H is None:
        (H, W) = frame.shape[:2]

    blob = cv2.dnn.blobFromImage(frame,1.0, (W,H),mean_value)
    net.setInput(blob)
    detections = net.forward()
    rects = []

    for i in range(0, detections.shape[2]):
        confidence = 0.5
        if detections[0, 0, i, 2] > confidence:
            box = detections[0,0,i,3:7] * np.array([W, H, W, H])
            rects.append(box.astype("int"))

            (startX, startY, endX, endY) = box.astype("int")
            cv2.rectangle(frame, (startX, startY), (endX, endY), (0, 255, 0),2)
    objects = centTrack.update(rects)

    for(objectID, centroid) in objects.items():

        text = "ID {}".format(objectID)
        cv2.putText(frame, text, (centroid[0] - 10, centroid[1] - 10),
                    cv2.FONT_HERSHEY_SIMPLEX,0.5, (0,255,0),2)
        cv2.circle(frame, (centroid[0],centroid[1]), 4, (0,255,0),-1)

    cv2.imshow("Test", frame)
    key = cv2.waitKey(1) & 0xFF
    if key == ord("q"):
        break

vs.release()
cv2.destroyAllWindows()

3.测试

单目标效果：

多目标效果：

缺点：

1. 视频的每帧运行一个对象检测器，计算量大，耗资源
2. 处理重叠目标对象效果差
3. 使用欧几里得距离，有时质心会出现交换ID情况