Identification, counting, and sizing of dispersed phase droplet of scanning electron microscopy micrograph using digital image processing

The identification of dispersed phase droplet in scanning electron microscopy (SEM) image is the heart of the process polymer blends, especially for the development of the polymeric materials and improvement of the polymer properties. Manual identification is the hard work and inaccurate method. To solve this problem, a digital image processing (DIP) method based on Hough transform is proposed for automatically identifies the dispersed phase droplet in SEM images. By combining the characteristics of SEM images and the DIP method, this method performed hierarchical Hough transform on the circular droplet to detect the object boundary in the SEM images. The DIP method has been experimented on variety of SEM images and very promising results have been achieved given more accuracy. Experimental results show that the proposed method with high adaptability is more accurate and rapidly than the traditional method.