5. Conclusion

In this work, we propose a new method for reconstructing a 3D vascular tree model from the human retina image by fractal interpolation. Firstly, we describe the method to extract the human retina vascular tree, and then we present a 3D reconstruction algorithm of the blood vessel. In the second step, we applied the Douglas-Peucker algorithm of simplification to detect control points and to compress the data in this 3D model. In the last step, the 3D fractal interpolation and control points are used to generate new data points and restore the original 3D model. From the obtained result, it is found that the Douglas-Peucker method has a high reduction ratio between 72 and 94% for the first test image and between 73 and 95% for the second image. According to the obtained results after the application of the proposed 3D fractal interpolation, we find that the error is small and ranging between 0.2 and 2% for the iteration number greater than 400 iterations. On the other hand, the error increases while decreasing the

iteration number. The advantage of this method is that there are more interpolated data points than the original data points; therefore, the reconstruction model is powerful and provides valuable information.
