Image of Wood with Defects Processing Based on Fractal Theory

It is important to detect defects in wood accurately for using forest resource efficiently. X-ray was used to test the defects based on the difference of X-ray intensity. To find out wood defects and their detail information of edges, fractal mathematical model was set up. According to fractal Brownian random field theory, in only tiny region of the picture, the surface of gray was self-similar in statistics. But for the edges, which the points located in the boundaries at adjacent region, this regular feature would be lost. The fractal parameters of these points were out of range. In this case we could finger out the edges of defects by calculating the fractal parameter H. We optimized the fractal algorithm, reduced calculating time for real-time request. The fractal algorithm was programmed finally for computer use. The experimental results show that the method is very successful for detecting wood defects. It is significant for wood manufacture to apply fractal theory on image processing and pattern recognition