{"id":262588,"date":"2024-05-15T14:39:19","date_gmt":"2024-05-15T06:39:19","guid":{"rendered":""},"modified":"2024-05-15T14:39:26","modified_gmt":"2024-05-15T06:39:26","slug":"kcf-%e8%b7%9f%e8%b8%aa%e7%ae%97%e6%b3%95%e4%bb%a3%e7%a0%81%e6%80%8e%e4%b9%88%e5%86%99","status":"publish","type":"post","link":"https:\/\/docs.pingcode.com\/ask\/262588.html","title":{"rendered":"kcf \u8ddf\u8e2a\u7b97\u6cd5\u4ee3\u7801\u600e\u4e48\u5199"},"content":{"rendered":"

\"kcf<\/p>\n

\u4e00\u3001\u6838\u5fc3\u539f\u7406\u89e3\u6790<\/h2>\n<\/p>\n

KCF\uff08Kernelized Correlation Filters\uff09\u8ddf\u8e2a\u7b97\u6cd5\u662f\u4e00\u79cd\u9ad8\u6548\u7684\u76ee\u6807\u8ddf\u8e2a\u6846\u67b6\uff0c\u4e3b\u8981\u901a\u8fc7\u5faa\u73af\u77e9\u9635\u548c\u5feb\u901f\u5085\u91cc\u53f6\u53d8\u6362\u6765\u5b9e\u73b0\u5feb\u901f\u8ddf\u8e2a\u3002\u8be5\u7b97\u6cd5\u7684\u6838\u5fc3\u4f18\u52bf\u5728\u4e8e\u901f\u5ea6\u5feb\u3001\u6027\u80fd\u597d\u3001\u5b9e\u73b0\u7b80\u5355<\/strong>\u3002\u5728\u5c55\u5f00\u8be6\u7ec6\u63cf\u8ff0\u4e4b\u524d\uff0c\u9996\u5148\u8981\u4e86\u89e3\u7b97\u6cd5\u7684\u5173\u952e\u6280\u672f\u70b9\uff1a\u4f7f\u7528\u5faa\u73af\u6837\u672c\u6765\u8bad\u7ec3\u76f8\u5173\u6ee4\u6ce2\u5668\uff0c\u5229\u7528\u6838\u6280\u5de7\u6765\u589e\u5f3a\u975e\u7ebf\u6027\u7279\u5f81\u8868\u8fbe\u80fd\u529b\uff0c\u4ee5\u53ca\u5229\u7528\u5feb\u901f\u5085\u91cc\u53f6\u53d8\u6362\uff08FFT\uff09\u5728\u9891\u57df\u8fdb\u884c\u9ad8\u6548\u7387\u7684\u8ba1\u7b97\u3002<\/p>\n<\/p>\n

\u5176\u4e2d\uff0c\u5faa\u73af\u77e9\u9635\u7684\u4f7f\u7528\u5141\u8bb8\u6211\u4eec\u7528DFT\uff08\u79bb\u6563\u5085\u91cc\u53f6\u53d8\u6362\uff09\u5c06\u5377\u79ef\u8fd0\u7b97\u8f6c\u6362\u4e3a\u5bf9\u5e94\u5143\u7d20\u7684\u4e58\u6cd5\u8fd0\u7b97\uff0c\u4ece\u800c\u964d\u4f4e\u4e86\u7b97\u6cd5\u7684\u590d\u6742\u5ea6\u3002\u6b64\u5916\uff0c\u901a\u8fc7\u5f15\u5165\u6838\u51fd\u6570\uff0c\u7b97\u6cd5\u53ef\u4ee5\u5904\u7406\u7ebf\u6027\u4e0d\u53ef\u5206\u7684\u6570\u636e\uff0c\u589e\u5f3a\u4e86\u8ddf\u8e2a\u7a33\u5b9a\u6027\u3002<\/p>\n<\/p>\n

\u4e8c\u3001\u73af\u5883\u642d\u5efa\u4e0e\u4f9d\u8d56<\/h2>\n<\/p>\n

\u5728\u7f16\u5199KCF\u8ddf\u8e2a\u7b97\u6cd5\u4ee3\u7801\u524d\uff0c\u9700\u8981\u51c6\u5907\u597d\u7f16\u7a0b\u73af\u5883\uff0c\u5e76\u4e14\u786e\u4fdd\u5b89\u88c5\u4e86\u5fc5\u8981\u7684\u5e93\u548c\u4f9d\u8d56\u3002\u901a\u5e38\uff0cKCF\u7b97\u6cd5\u53ef\u5728Python\u7b49\u7f16\u7a0b\u8bed\u8a00\u4e2d\u5b9e\u73b0\uff0c\u800cPython\u7684OpenCV\u5e93\u5c31\u63d0\u4f9b\u4e86\u4e30\u5bcc\u7684\u56fe\u50cf\u5904\u7406\u529f\u80fd\uff0c\u5bf9\u4e8e\u7b97\u6cd5\u5b9e\u73b0\u81f3\u5173\u91cd\u8981\u3002<\/p>\n<\/p>\n

\u9996\u5148\u9700\u8981\u786e\u4fdd\u5df2\u7ecf\u5b89\u88c5Python\u73af\u5883\uff0c\u7136\u540e\u901a\u8fc7pip\u5b89\u88c5OpenCV\u5e93\uff1a<\/p>\n<\/p>\n

pip install opencv-python<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u6b64\u5916\uff0c\u4e3a\u4e86\u5b8c\u6210KCF\u7b97\u6cd5\u7684\u5b9e\u73b0\uff0c\u60a8\u53ef\u80fd\u8fd8\u9700\u8981Numpy\u5e93\u8fdb\u884c\u77e9\u9635\u7684\u5feb\u901f\u64cd\u4f5c\u3002\u5982\u679c\u5c1a\u672a\u5b89\u88c5\uff0c\u53ef\u4ee5\u901a\u8fc7\u4ee5\u4e0b\u547d\u4ee4\u8fdb\u884c\u5b89\u88c5\uff1a<\/p>\n<\/p>\n
pip install numpy<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u4e09\u3001\u7b97\u6cd5\u5b9e\u73b0\u6b65\u9aa4<\/h2>\n<\/p>\n
KCF\u7b97\u6cd5\u7684\u5b9e\u73b0\u53ef\u4ee5\u5206\u4e3a\u6a21\u578b\u521d\u59cb\u5316\u3001\u6a21\u578b\u8bad\u7ec3\u3001\u76ee\u6807\u68c0\u6d4b\u4e0e\u66f4\u65b0<\/strong>\u51e0\u4e2a\u4e3b\u8981\u6b65\u9aa4\u3002<\/p>\n<\/p>\n
\u6a21\u578b\u521d\u59cb\u5316<\/h3>\n<\/p>\n
\u5728\u8ddf\u8e2a\u5f00\u59cb\u4e4b\u524d\uff0c\u9996\u5148\u9700\u8981\u5b9a\u4e49\u4e00\u4e2a\u7528\u4e8e\u8ddf\u8e2a\u7684\u533a\u57df\uff08\u5373\u521d\u59cb\u76ee\u6807\u6846\uff09\uff0c\u5e76\u63d0\u53d6\u8be5\u6846\u5185\u7684\u7279\u5f81\u7528\u4e8e\u521d\u59cb\u5316\u6a21\u578b\u3002\u901a\u5e38\u4f7f\u7528HOG\uff08\u65b9\u5411\u68af\u5ea6\u76f4\u65b9\u56fe\uff09\u4f5c\u4e3a\u7279\u5f81\u63cf\u8ff0\u5b50\u56e0\u5176\u5728\u76ee\u6807\u8ddf\u8e2a\u4e2d\u7684\u8868\u73b0\u826f\u597d\u3002\u9700\u8981\u8ba1\u7b97\u51fa\u521d\u59cb\u76ee\u6807\u7684HOG\u7279\u5f81\uff0c\u5e76\u4f7f\u7528\u8fd9\u4e9b\u7279\u5f81\u6765\u8bad\u7ec3\u7b2c\u4e00\u4e2aKCF\u6a21\u578b\u3002<\/p>\n<\/p>\n
\u6a21\u578b\u8bad\u7ec3<\/h3>\n<\/p>\n
KCF\u6a21\u578b\u7684\u8bad\u7ec3\u901a\u5e38\u5305\u62ec\u751f\u6210\u4e00\u4e2a\u5faa\u73af\u77e9\u9635\uff0c\u5e76\u901a\u8fc7\u4e00\u4e2a\u6838\u51fd\u6570\u5bf9\u7279\u5f81\u8fdb\u884c\u6620\u5c04\u3002\u7136\u540e\u5728\u9891\u57df\u4e2d\u4f7f\u7528\u5feb\u901f\u5085\u91cc\u53f6\u53d8\u6362\u5bf9\u6a21\u578b\u8fdb\u884c\u8bad\u7ec3\u3002\u5728\u6b64\u6b65\u9aa4\u4e2d\uff0c\u4e3b\u8981\u5de5\u4f5c\u662f\u8ba1\u7b97\u76f8\u5173\u6ee4\u6ce2\u5668\u5e76\u4f18\u5316\u5176\u53c2\u6570\u3002<\/p>\n<\/p>\n
\u76ee\u6807\u68c0\u6d4b\u4e0e\u66f4\u65b0<\/h3>\n<\/p>\n
\u5728\u6bcf\u4e00\u5e27\u89c6\u9891\u4e2d\uff0c\u7b97\u6cd5\u9700\u8981\u6839\u636e\u4e0a\u4e00\u5e27\u7684\u6a21\u578b\u6765\u9884\u6d4b\u76ee\u6807\u7684\u65b0\u4f4d\u7f6e\uff0c\u5e76\u66f4\u65b0\u6a21\u578b\u3002\u5728\u9884\u6d4b\u6b65\u9aa4\u4e2d\uff0c\u7b97\u6cd5\u9700\u8981\u5bf9\u65b0\u5e27\u4e2d\u7684\u76ee\u6807\u533a\u57df\u63d0\u53d6\u7279\u5f81\uff0c\u5e76\u5229\u7528\u8bad\u7ec3\u597d\u7684KCF\u6a21\u578b\u8fdb\u884c\u54cd\u5e94\u56fe\u7684\u751f\u6210\uff0c\u6700\u5927\u54cd\u5e94\u503c\u7684\u4f4d\u7f6e\u5373\u4e3a\u76ee\u6807\u7684\u65b0\u4f4d\u7f6e\u3002\u5728\u66f4\u65b0\u6b65\u9aa4\u4e2d\uff0c\u5c06\u4f1a\u7528\u65b0\u83b7\u5f97\u7684\u76ee\u6807\u7279\u5f81\u6765\u66f4\u65b0KCF\u6a21\u578b\uff0c\u4ee5\u4fbf\u4e8e\u8fdb\u884c\u4e0b\u4e00\u6b21\u8fed\u4ee3\u68c0\u6d4b\u3002<\/p>\n<\/p>\n
\u56db\u3001\u7b97\u6cd5\u7ec6\u8282\u4e0e\u4ee3\u7801\u5b9e\u73b0<\/h2>\n<\/p>\n
\u63a5\u4e0b\u6765\u5c06\u5177\u4f53\u5b9e\u73b0KCF\u7b97\u6cd5\u7684\u6838\u5fc3\u4ee3\u7801\u3002\u9700\u8981\u6ce8\u610f\u7684\u662f\uff0c\u8fd9\u91cc\u4ec5\u63d0\u4f9b\u4f7f\u7528Python\u8bed\u8a00\u5b9e\u73b0\u7684\u57fa\u672c\u6846\u67b6\uff0c\u5b9e\u9645\u6548\u679c\u4f1a\u53d7\u5230\u53c2\u6570\u9009\u62e9\u3001\u7279\u5f81\u63d0\u53d6\u65b9\u6cd5\u7b49\u591a\u79cd\u56e0\u7d20\u7684\u5f71\u54cd\u3002<\/p>\n<\/p>\n
\u9996\u5148\u662f\u63d0\u53d6HOG\u7279\u5f81\u7684\u51fd\u6570\uff0c\u56e0\u4f7f\u7528OpenCV\u5e93\uff0c\u53ef\u76f4\u63a5\u8c03\u7528\u76f8\u5173\u51fd\u6570\u83b7\u53d6\u7279\u5f81\uff1a<\/p>\n<\/p>\n
import cv2<\/p>\n
import numpy as np<\/p>\n
def extract_hog_features(image, bbox):<\/p>\n
    # \u83b7\u53d6HOG\u7279\u5f81\u63cf\u8ff0\u5b50<\/p>\n
    hog = cv2.HOGDescriptor()<\/p>\n
    # \u63d0\u53d6\u76ee\u6807\u533a\u57df\u7684HOG\u7279\u5f81<\/p>\n
    x, y, w, h = bbox<\/p>\n
    target_region = image[y:y+h, x:x+w]<\/p>\n
    # \u9002\u5f53\u8c03\u6574\u5927\u5c0f\u4ee5\u5339\u914d\u9884\u8bbe\u7684\u7a97\u53e3\u5c3a\u5bf8<\/p>\n
    target_region = cv2.resize(target_region, (64, 128))<\/p>\n
    hog_features = hog.compute(target_region)<\/p>\n
    return hog_features<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u7d27\u968f\u5176\u540e\uff0c\u662f\u521b\u5efa\u5faa\u73af\u77e9\u9635\u4e0e\u8bad\u7ec3\u76f8\u5173\u6ee4\u6ce2\u5668\u7684\u4ee3\u7801\uff1a<\/p>\n<\/p>\n
def create_circular_matrix(features):<\/p>\n
    # \u4f7f\u7528\u5085\u91cc\u53f6\u53d8\u6362\u521b\u5efa\u5faa\u73af\u77e9\u9635<\/p>\n
    dft_features = np.fft.fft2(features)<\/p>\n
    circular_matrix = np.real(np.fft.ifft2(np.multiply(dft_features, np.conj(dft_features))))<\/p>\n
    return circular_matrix<\/p>\n
def trAI<\/a>n_kcf_model(hog_features, target_size):<\/p>\n
    # \u6839\u636eHOG\u7279\u5f81\u4e0e\u76ee\u6807\u5c3a\u5bf8\u8bad\u7ec3KCF\u6a21\u578b<\/p>\n
    # \u6b64\u5904\u7701\u7565\u5177\u4f53\u7ec6\u8282\uff0c\u5b9e\u9645\u4ee3\u7801\u6839\u636e\u7b97\u6cd5\u516c\u5f0f\u5b9e\u73b0<\/p>\n
    pass<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u5bf9\u4e8e\u76ee\u6807\u7684\u68c0\u6d4b\u4e0e\u6a21\u578b\u66f4\u65b0\uff0c\u6d41\u7a0b\u5982\u4e0b\uff1a<\/p>\n<\/p>\n
def detect_target(new_frame, kcf_model):<\/p>\n
    # \u5728\u65b0\u5e27\u4e2d\u68c0\u6d4b\u76ee\u6807\uff0c\u8fd4\u56de\u76ee\u6807\u7684\u65b0\u4f4d\u7f6e<\/p>\n
    # \u6b64\u5904\u7701\u7565\u5177\u4f53\u5b9e\u73b0<\/p>\n
    pass<\/p>\n
def update_kcf_model(kcf_model, new_features):<\/p>\n
    # \u4f7f\u7528\u65b0\u7279\u5f81\u66f4\u65b0KCF\u6a21\u578b<\/p>\n
    # \u6b64\u5904\u7701\u7565\u5177\u4f53\u5b9e\u73b0<\/p>\n
    pass<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u6700\u540e\uff0c\u662f\u5c06\u4ee5\u4e0a\u90e8\u5206\u7ec4\u5408\u8d77\u6765\u5b8c\u6210\u89c6\u9891\u4e2d\u76ee\u6807\u7684\u8fde\u7eed\u8ddf\u8e2a\uff1a<\/p>\n<\/p>\n
video_path = 'path\/to\/your\/video'<\/p>\n
cap = cv2.VideoCapture(video_path)<\/p>\n
\u8bfb\u53d6\u89c6\u9891\u7684\u7b2c\u4e00\u5e27\u5e76\u9009\u62e9\u76ee\u6807\u533a\u57df<\/strong><\/h2>\n
ret, first_frame = cap.read()<\/p>\n
bbox = cv2.selectROI('Select Target', first_frame, False)<\/p>\n
cv2.destroyAllWindows()<\/p>\n
\u521d\u59cb\u76ee\u6807\u7279\u5f81\u63d0\u53d6\u4e0e\u6a21\u578b\u8bad\u7ec3<\/strong><\/h2>\n
initial_features = extract_hog_features(first_frame, bbox)<\/p>\n
kcf_model = train_kcf_model(initial_features, bbox[2:4])<\/p>\n
while True:<\/p>\n
    # \u5faa\u73af\u8bfb\u53d6\u89c6\u9891\u5e27<\/p>\n
    ret, frame = cap.read()<\/p>\n
    if not ret:<\/p>\n
        break<\/p>\n
    # \u76ee\u6807\u68c0\u6d4b\u4e0e\u6a21\u578b\u66f4\u65b0<\/p>\n
    target_position = detect_target(frame, kcf_model)<\/p>\n
    new_features = extract_hog_features(frame, target_position)<\/p>\n
    update_kcf_model(kcf_model, new_features)<\/p>\n
    # \u8ddf\u8e2a\u7ed3\u679c\u7684\u53ef\u89c6\u5316<\/p>\n
    p1 = (int(target_position[0]), int(target_position[1]))<\/p>\n
    p2 = (int(target_position[0] + target_position[2]), int(target_position[1] + target_position[3]))<\/p>\n
    cv2.rectangle(frame, p1, p2, (255,0,0), 2, 1)<\/p>\n
    cv2.imshow('KCF Tracker', frame)<\/p>\n
    if cv2.waitKey(100) & 0xFF == ord('q'):<\/p>\n
        break<\/p>\n
cap.release()<\/p>\n
cv2.destroyAllWindows()<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u5728\u7f16\u5199\u5b8c\u6574\u7684KCF\u8ddf\u8e2a\u7b97\u6cd5\u4ee3\u7801\u65f6\uff0c\u6bcf\u4e00\u6b65\u90fd\u9700\u8981\u7ec6\u81f4\u7684\u6570\u5b66\u63a8\u5bfc\u5e76\u8003\u8651\u8ba1\u7b97\u7684\u6548\u7387\u3002\u4e0a\u8ff0\u4ee3\u7801\u53ea\u63d0\u4f9b\u4e86\u4e00\u4e2a\u6846\u67b6\u53c2\u8003\uff0c\u5e76\u672a\u5305\u542b\u5168\u90e8\u7684\u5b9e\u73b0\u7ec6\u8282\u3002\u5b8c\u6574\u548c\u4f18\u5316\u7684KCF\u7b97\u6cd5\u5b9e\u73b0\u8fd8\u9700\u8981\u8003\u8651\u76ee\u6807\u5c3a\u5ea6\u53d8\u5316\u3001\u65cb\u8f6c\u4e0d\u53d8\u6027\u7b49\u95ee\u9898\uff0c\u5e76\u901a\u8fc7\u5927\u91cf\u5b9e\u9a8c\u8fdb\u884c\u8c03\u53c2\u4ee5\u786e\u4fdd\u8ddf\u8e2a\u7684\u51c6\u786e\u6027\u548c\u9c81\u68d2\u6027\u3002\u5728\u5b9e\u9645\u4f7f\u7528\u4e2d\uff0c\u53ef\u4ee5\u53c2\u7167\u5f00\u6e90\u793e\u533a\u4e2d\u6210\u719f\u7684KCF\u7b97\u6cd5\u5b9e\u73b0\u8fdb\u884c\u7814\u7a76\u548c\u5f00\u53d1\u3002<\/p>\n<\/p>\n
\u76f8\u5173\u95ee\u7b54FAQs\uff1a<\/strong><\/h2>\n
    \n
  1. \n
    \u5982\u4f55\u7f16\u5199KCF\u8ddf\u8e2a\u7b97\u6cd5\u7684\u4ee3\u7801\uff1f<\/strong>
    \nKCF\uff08Kernelized Correlation Filters\uff09\u8ddf\u8e2a\u7b97\u6cd5\u662f\u4e00\u79cd\u57fa\u4e8e\u76f8\u5173\u6ee4\u6ce2\u5668\u7684\u76ee\u6807\u8ddf\u8e2a\u7b97\u6cd5\u3002\u7f16\u5199KCF\u8ddf\u8e2a\u7b97\u6cd5\u7684\u4ee3\u7801\u9700\u8981\u638c\u63e1\u76f8\u5173\u6ee4\u6ce2\u5668\u7684\u539f\u7406\u548c\u5b9e\u73b0\u65b9\u6cd5\u3002\u4f60\u53ef\u4ee5\u5148\u4e86\u89e3\u76f8\u5173\u6ee4\u6ce2\u5668\u7684\u57fa\u672c\u7406\u8bba\u4ee5\u53ca\u5728\u76ee\u6807\u8ddf\u8e2a\u4e2d\u7684\u5e94\u7528\u3002\u7136\u540e\uff0c\u4f60\u53ef\u4ee5\u6309\u7167\u4ee5\u4e0b\u6b65\u9aa4\u7f16\u5199KCF\u8ddf\u8e2a\u7b97\u6cd5\u7684\u4ee3\u7801\uff1a<\/p>\n<\/li>\n
  2. \n
    \u6570\u636e\u9884\u5904\u7406\uff1a\u8bfb\u53d6\u89c6\u9891\u5e27\uff0c\u5bf9\u5e27\u8fdb\u884c\u7070\u5ea6\u8f6c\u6362\u3001\u5c3a\u5bf8\u8c03\u6574\u7b49\u9884\u5904\u7406\u64cd\u4f5c\uff0c\u4ee5\u63d0\u9ad8\u7b97\u6cd5\u7684\u6548\u7387\u3002<\/p>\n<\/li>\n
  3. \n
    \u83b7\u5f97\u76ee\u6807\u6a21\u677f\uff1a\u5728\u7b2c\u4e00\u5e27\u4e2d\u624b\u52a8\u9009\u62e9\u611f\u5174\u8da3\u7684\u76ee\u6807\u533a\u57df\uff0c\u5e76\u5c06\u76ee\u6807\u533a\u57df\u4f5c\u4e3a\u521d\u59cb\u6a21\u677f\u3002<\/p>\n<\/li>\n
  4. \n
    \u7279\u5f81\u63d0\u53d6\uff1a\u5bf9\u6a21\u677f\u533a\u57df\u63d0\u53d6\u7279\u5f81\uff0c\u5e38\u7528\u7684\u7279\u5f81\u5305\u62ecHOG\u7279\u5f81\u3001\u989c\u8272\u76f4\u65b9\u56fe\u7b49\u3002<\/p>\n<\/li>\n
  5. \n
    \u521d\u59cb\u5316\u6ee4\u6ce2\u5668\uff1a\u4f7f\u7528\u63d0\u53d6\u7684\u7279\u5f81\u521d\u59cb\u5316\u76f8\u5173\u6ee4\u6ce2\u5668\uff0c\u5305\u62ec\u8bad\u7ec3\u6ee4\u6ce2\u5668\u3001\u8ba1\u7b97\u54cd\u5e94\u7b49\u6b65\u9aa4\u3002<\/p>\n<\/li>\n
  6. \n
    \u76ee\u6807\u8ddf\u8e2a\uff1a\u5bf9\u4e8e\u540e\u7eed\u5e27\uff0c\u6839\u636e\u4e0a\u4e00\u5e27\u7684\u76ee\u6807\u4f4d\u7f6e\u548c\u9884\u6d4b\u7ed3\u679c\uff0c\u901a\u8fc7\u6ee4\u6ce2\u5668\u5bf9\u65b0\u5e27\u8fdb\u884c\u8ddf\u8e2a\uff0c\u5f97\u5230\u76ee\u6807\u7684\u65b0\u4f4d\u7f6e\u3002<\/p>\n<\/li>\n
  7. \n
    \u66f4\u65b0\u6ee4\u6ce2\u5668\uff1a\u6839\u636e\u65b0\u5e27\u7684\u9884\u6d4b\u7ed3\u679c\uff0c\u4f7f\u7528\u5728\u7ebf\u5b66\u4e60\u65b9\u6cd5\u66f4\u65b0\u6ee4\u6ce2\u5668\uff0c\u4ee5\u9002\u5e94\u76ee\u6807\u5916\u89c2\u7684\u53d8\u5316\u3002<\/p>\n<\/li>\n
  8. \n
    \u8f93\u51fa\u8ddf\u8e2a\u7ed3\u679c\uff1a\u5c06\u8ddf\u8e2a\u7ed3\u679c\u8f93\u51fa\uff0c\u53ef\u4ee5\u901a\u8fc7\u7ed8\u5236\u8fb9\u754c\u6846\u3001\u6807\u8bb0\u76ee\u6807\u7b49\u65b9\u5f0f\u5c55\u793a\u8ddf\u8e2a\u6548\u679c\u3002<\/p>\n<\/li>\n<\/ol>\n
    \u7f16\u5199KCF\u8ddf\u8e2a\u7b97\u6cd5\u7684\u4ee3\u7801\u9700\u8981\u638c\u63e1\u76f8\u5173\u7684\u7f16\u7a0b\u8bed\u8a00\uff0c\u5982Python\u3001C++\u7b49\uff0c\u5e76\u501f\u52a9\u4e00\u4e9b\u5f00\u6e90\u8ddf\u8e2a\u7b97\u6cd5\u5e93\uff0c\u5982OpenCV\u3001MATLAB\u7b49\u6765\u52a0\u901f\u5f00\u53d1\u8fc7\u7a0b\u3002\u540c\u65f6\uff0c\u4e86\u89e3KCF\u7b97\u6cd5\u7684\u4f18\u7f3a\u70b9\u548c\u5e94\u7528\u573a\u666f\u4e5f\u80fd\u5e2e\u52a9\u4f60\u66f4\u597d\u5730\u7f16\u5199\u4ee3\u7801\u3002<\/p>\n
      \n
    1. KCF\u8ddf\u8e2a\u7b97\u6cd5\u7684\u4ee3\u7801\u793a\u4f8b\u662f\u4ec0\u4e48\uff1f<\/strong>
      \n\u4ee5\u4e0b\u662f\u4e00\u4e2a\u7b80\u5355\u7684KCF\u8ddf\u8e2a\u7b97\u6cd5\u7684\u4ee3\u7801\u793a\u4f8b\uff1a<\/li>\n<\/ol>\n
      import cv2\ntracker = cv2.TrackerKCF_create()  # \u521b\u5efaKCF\u8ddf\u8e2a\u5668\n\nvideo = cv2.VideoCapture("video.mp4")  # \u6253\u5f00\u89c6\u9891\u6587\u4ef6\n\nsuccess, frame = video.read()  # \u8bfb\u53d6\u7b2c\u4e00\u5e27\nbbox = cv2.selectROI("Tracking", frame, False)  # \u9009\u62e9\u76ee\u6807\u533a\u57df\ntracker.init(frame, bbox)  # \u521d\u59cb\u5316\u8ddf\u8e2a\u5668\n\nwhile True:\n    success, frame = video.read()  # \u8bfb\u53d6\u65b0\u5e27\n    if not success:\n        break\n    success, bbox = tracker.update(frame)  # \u66f4\u65b0\u76ee\u6807\u4f4d\u7f6e\n    if success:\n        x, y, w, h = [int(i) for i in bbox]\n        cv2.rectangle(frame, (x, y), (x + w, y + h), (0, 255, 0), 2)  # \u7ed8\u5236\u8fb9\u754c\u6846\n    cv2.imshow("Tracking", frame)  # \u663e\u793a\u8ddf\u8e2a\u7ed3\u679c\n    if cv2.waitKey(1) == ord('q'):\n        break\n\nvideo.release()  # \u91ca\u653e\u89c6\u9891\u6587\u4ef6\ncv2.destroyAllWindows()  # \u5173\u95ed\u7a97\u53e3\n<\/code><\/pre>\n
      \u8fd9\u6bb5\u4ee3\u7801\u4f7f\u7528OpenCV\u5e93\u63d0\u4f9b\u7684KCF\u8ddf\u8e2a\u5668\uff0c\u5728\u7ed9\u5b9a\u7684\u89c6\u9891\u6587\u4ef6\u4e2d\u9009\u62e9\u76ee\u6807\u533a\u57df\uff0c\u5e76\u5b9e\u65f6\u8ddf\u8e2a\u76ee\u6807\u7684\u4f4d\u7f6e\u3002\u901a\u8fc7\u8c03\u7528tracker.update(frame)<\/code>\u65b9\u6cd5\u6765\u66f4\u65b0\u76ee\u6807\u4f4d\u7f6e\uff0c\u5e76\u7528\u77e9\u5f62\u6846\u5c06\u76ee\u6807\u533a\u57df\u6807\u8bb0\u51fa\u6765\u3002\u6700\u540e\uff0c\u901a\u8fc7cv2.imshow()<\/code>\u65b9\u6cd5\u5c06\u8ddf\u8e2a\u7ed3\u679c\u663e\u793a\u51fa\u6765\u3002<\/p>\n
        \n
      1. KCF\u8ddf\u8e2a\u7b97\u6cd5\u9002\u5408\u7528\u4e8e\u54ea\u4e9b\u573a\u666f\uff1f<\/strong>
        \nKCF\u8ddf\u8e2a\u7b97\u6cd5\u9002\u7528\u4e8e\u5404\u79cd\u76ee\u6807\u8ddf\u8e2a\u573a\u666f\uff0c\u5305\u62ec\u4f46\u4e0d\u9650\u4e8e\u4ee5\u4e0b\u51e0\u79cd\uff1a<\/li>\n<\/ol>\n
          \n
        • \u81ea\u52a8\u9a7e\u9a76\uff1a\u5728\u81ea\u52a8\u9a7e\u9a76\u9886\u57df\uff0c\u53ef\u4ee5\u5229\u7528KCF\u7b97\u6cd5\u5b9e\u73b0\u8f66\u8f86\u3001\u884c\u4eba\u3001\u4ea4\u901a\u6807\u5fd7\u7b49\u76ee\u6807\u7684\u5b9e\u65f6\u8ddf\u8e2a\uff0c\u4ee5\u63d0\u4f9b\u73af\u5883\u611f\u77e5\u548c\u51b3\u7b56\u652f\u6301\u3002<\/li>\n
        • \u89c6\u9891\u76d1\u63a7\uff1a\u5728\u89c6\u9891\u76d1\u63a7\u573a\u666f\u4e2d\uff0c\u53ef\u4ee5\u5229\u7528KCF\u7b97\u6cd5\u5bf9\u4eba\u5458\u3001\u8f66\u8f86\u7b49\u76ee\u6807\u8fdb\u884c\u8ddf\u8e2a\uff0c\u4ee5\u5b9e\u73b0\u5b9e\u65f6\u76d1\u63a7\u548c\u4e8b\u4ef6\u68c0\u6d4b\u3002<\/li>\n
        • \u65e0\u4eba\u673a\u8ddf\u8e2a\uff1aKCF\u7b97\u6cd5\u53ef\u4ee5\u88ab\u5e94\u7528\u4e8e\u65e0\u4eba\u673a\u7684\u76ee\u6807\u8ddf\u8e2a\u4efb\u52a1\uff0c\u5b9e\u65f6\u8ffd\u8e2a\u8239\u53ea\u3001\u8f66\u8f86\u3001\u884c\u4eba\u7b49\u76ee\u6807\uff0c\u4ece\u800c\u5b9e\u73b0\u65e0\u4eba\u673a\u7684\u667a\u80fd\u81ea\u4e3b\u98de\u884c\u3002<\/li>\n
        • \u589e\u5f3a\u73b0\u5b9e\uff1a\u5728\u589e\u5f3a\u73b0\u5b9e\u5e94\u7528\u4e2d\uff0cKCF\u7b97\u6cd5\u53ef\u4ee5\u7528\u4e8e\u76ee\u6807\u7684\u5b9e\u65f6\u8ddf\u8e2a\u548c\u59ff\u6001\u4f30\u8ba1\uff0c\u63d0\u4f9b\u66f4\u771f\u5b9e\u3001\u66f4\u7a33\u5b9a\u7684\u865a\u5b9e\u878d\u5408\u6548\u679c\u3002<\/li>\n<\/ul>\n
          \u603b\u7684\u6765\u8bf4\uff0cKCF\u8ddf\u8e2a\u7b97\u6cd5\u5177\u6709\u8f83\u9ad8\u7684\u5b9e\u65f6\u6027\u548c\u9c81\u68d2\u6027\uff0c\u9002\u7528\u4e8e\u5404\u79cd\u9700\u8981\u5b9e\u65f6\u76ee\u6807\u8ddf\u8e2a\u7684\u573a\u666f\u3002\u901a\u8fc7\u5408\u7406\u7684\u53c2\u6570\u8bbe\u7f6e\u548c\u4f18\u5316\uff0c\u53ef\u4ee5\u63d0\u9ad8\u7b97\u6cd5\u7684\u8ddf\u8e2a\u6548\u679c\u548c\u6027\u80fd\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"\u4e00\u3001\u6838\u5fc3\u539f\u7406\u89e3\u6790 KCF\uff08Kernelized Correlation Filters\uff09\u8ddf\u8e2a\u7b97\u6cd5\u662f\u4e00\u79cd\u9ad8\u6548\u7684\u76ee […]","protected":false},"author":3,"featured_media":262595,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[37],"tags":[],"acf":[],"_links":{"self":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/262588"}],"collection":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/comments?post=262588"}],"version-history":[{"count":0,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/262588\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/media\/262595"}],"wp:attachment":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/media?parent=262588"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/categories?post=262588"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/tags?post=262588"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}