Obtaining the porosity of porous silicon from sem images based on the image processing method

This project presents a new method to characterize porous silicon (PS) from the scanning electron microscope (SEM) image. Conventionally, the porosity from the PS film is able to be determined by using gravimetric or quasigravimetric method. However, the gravimetric method requires various stages of processing to obtain the porosity of PS where using the gravimetric method, the mass of the silicon sample is measured despite the fact that the final result would suggest that the porous film would be destructed. This is due to the usage of an alkaline solution that dissolves the sample. In addition, the problem of this method is that it will provide low accuracy of the porosity characterization. In this paper, the new method of characterization which uses the image processing technique is proposed. It uses segmentation via a clustering algorithm called the AFKM. The objective is to determine the porosity of PS with a non-destructive manner which will not harm the sample. The implementation of the AFKM technique as image segmentation is to segment the image into two significant regions i.e. the holes (black region) and the background (white region) of the image so that it can easily be identified. Next, the area for the black region is calculated in order to obtain the porosity of the PS. The porosity is calculated by calculating the ratio of the black area to the total area of SEM images. From the porosity calculation, their stability can be identified. Furthermore, this method provides a more accurate determination of the total area of the pore sizes and most importantly, it will not destruct the sample