A dataset of micro-CT images are used. The denoising filter is used to remove noise from the image as a pre-processing step, then feature extraction is performed, then the back propagation neural network is created, and finally, it changes the number weight of the network and records the output. The proposed algorithm is compared with the thresholding method and the region growing method. The results showed that the proposed technique outperforms other methods in terms of speed and accuracy..