Unsupervised Scoliosis Diagnosis via a Joint Recognition Method with Multifeature Descriptors and Centroids Extraction
Algorithm 1
The training process of the cascade gentle AdaBoost classifier.
Input: the cascade layer number K, the minimum detection rate d (D0 = 1), the maximum false-positive rate f (F0 = 1), the positive sample set P, the negative sample set N, and the final false-positive rate Ft.
Output: subwindows of the image belonging to positive samples
(1)
for i = 0:K, Fi > Ftdo
(2)
++i; ni = 0; Fi = Fi−1;
(3)
if Fi > f × Fi−1then
(4)
++ni; train strong classifier with ni weak classifier using AdaBoost on P and N;
(5)
Calculate the detection rate Di and false-positive rate Fi of the current classifier;
(6)
if di < d × Di−1then
(7)
Reduce the threshold of the ith layer strong classifier;
(8)
Calculate detection rate Di and false-positive rate Fi of the current classifier;
(9)
end if
(10)
end if
(11)
N = Φ; use the current classifier to detect testing images and put the negative image into N; output and number subwindows belonging to positive samples.
