Otsu-Kmeans gravity-based multi-spots center extraction method for microlens array imaging system

We propose an Otsu-Kmeans gravity-based center extraction method for multi-spot images captured by the microlens array imaging system. We develop a 4-connected-region-based region of interest location method to recognize and segment light spot areas. We design an Otsu-Kmeans gravity-based threshold selection algorithm to realize the multi-spot center accurate position. Both numerical and actual experiments are implemented. The effectiveness and feasibility of the proposed method are validated by comparing with the results from the traditional centroid method, the Hough circle method, the Gaussian fitting method, and the Otsu thresholding centroid method. The results show that the proposed method can realize high-precision positioning and can be used for the sub-image segmentation of the microlens array imaging system.