SK-FMYOLOV3: A Novel Detection Method for Urine Test Strips

Algorithm 1

The cluster box predicts flow that by fusing the fuzzy C-means algorithm into the K-means algorithm.

Input: image , classification

Step0: Standardize the input data, Eg. (3)(4), constrain the input value to [0,1].

Step1: Define the objective function and constraints of the algorithm, Eg. (5).

Step2: Obtain the cluster center and membership of the i pixel by the Langer multiplier method.

Step3: Use = (, , , ..., ) as the initial cluster center.

Step4: Calculate the distance from each sample to the cluster center according to the Eg. (6), (7).

Step5: Divide the sample into the class corresponding to the cluster center with the smallest distance.

Step6: Update the cluster center by the Eg. (8).

Step7: Repeat Step5 and Step6 until the value of is unchanged.

Output: output prediction box center