如何在cv2.HOGDescriptor中使用自定义SVM检测器？遇尺寸断言错误

问题背景 #

我正在学习HOG相关内容，尝试用自定义训练的SVM目标检测器替代cv2.HOGDescriptor_getDefaultPeopleDetector()。官方示例代码如下：

hog = cv2.HOGDescriptor()
hog.setSVMDetector(cv2.HOGDescriptor_getDefaultPeopleDetector())

但在应用自定义检测器时遇到了问题。

训练阶段代码 #

# 即使使用默认参数（不传入任何参数），问题仍然存在
hog = cv2.HOGDescriptor(_winSize = (64, 64), 
                        _blockSize = (16, 16), 
                        _blockStride = (2, 2),
                        _cellSize = (8, 8),
                        _nbins = 9)

# 所有正负样本图像均为裁剪后的边界框，已调整大小为(128, 128)
positive_features = np.array([hog.compute(img) for img in positive_images])
negative_features = np.array([hog.compute(img) for img in negative_images])

feature_matrix = np.concatenate((positive_features, negative_features), axis=0)
labels = np.concatenate((np.zeros(len(positive_features)), np.ones(len(negative_features))), axis=0) # 这违反直觉，但OpenCV SVM配合cv2.detectMultiScale()使用时，似乎期望正类为0，负类为1

train_matrix = np.concatenate((feature_matrix, np.expand_dims(labels, 0).T), axis = 1)

np.random.seed(int(SEED))
np.random.shuffle(train_matrix)

feature_matrix = train_matrix[:, :-1]
labels = train_matrix[:, -1]

# OpenCV格式处理
feature_matrix = feature_matrix.astype(np.float32)  # 转换为32位浮点数
labels = labels.astype(np.int32)  # 将标签转换为32位有符号整数

model = cv2.ml.SVM_create()
model.setType(cv2.ml.SVM_C_SVC)
model.setKernel(cv2.ml.SVM_LINEAR)
model.setTermCriteria((cv2.TERM_CRITERIA_MAX_ITER, 100, 1e-6))

model.train(feature_matrix, cv2.ml.ROW_SAMPLE, labels)
model.save("model.svm")

预测阶段代码 #

# 即使使用默认参数（不传入任何参数），问题仍然存在
hog = cv2.HOGDescriptor(_winSize = (64, 64), 
                        _blockSize = (16, 16), 
                        _blockStride = (2, 2),
                        _cellSize = (8, 8),
                        _nbins = 9)
model = cv2.ml.SVM_load("model.svm")
support_vectors = model.getSupportVectors()
coefficients = -model.getDecisionFunction(0)[0]

coefficients = np.array(coefficients).reshape(1, -1)
svmdetector = np.concatenate((support_vectors, coefficients), axis=1)
hog.setSVMDetector(svmdetector.T.flatten())

报错信息 #

hog.setSVMDetector(svmdetector.T.flatten())
cv2.error: OpenCV(4.7.0) /io/opencv/modules/objdetect/src/hog.cpp:120: error: (-215:Assertion failed) checkDetectorSize() in function 'setSVMDetector'

查看OpenCV源码，错误来自checkDetectorSize()函数，其定义如下：

bool HOGDescriptor::checkDetectorSize() const
{
    size_t detectorSize = svmDetector.size(), descriptorSize = getDescriptorSize();
    return detectorSize == 0 ||
        detectorSize == descriptorSize ||
        detectorSize == descriptorSize + 1;
}

另外还发现一个奇怪现象：

hog.setSVMDetector(cv2.HOGDescriptor_getDefaultPeopleDetector()) # 正常工作 [形状: (3781,)]
hog.setSVMDetector(cv2.HOGDescriptor_getDaimlerPeopleDetector()) # 无法工作 [形状: (1981,)]

内容的提问来源于stack exchange，提问作者hafiz031