An Image Segmentation Algorithm Based On Otsu Optimized By Fractional-order Particle Swarm Optimization

Image segmentation is the key step of image processing.It has an important influence on the edge object extraction of the image.Otsu segmentation method is often used in recent years,and it has a wide range of applications.But the Otsu algorithm has a long running time,an disadvantage of large amount of calculation.While traditional particle swarm optimization algorithm for optimizing Otsu algorithm has the disadvantages of slow convergence and easy to fall into local optimum.In view of the above,a fractional order calculus algorithm is proposed to optimize the particle swarm optimization algorithm.Combined with Otsu algorithm,fractional calculus algorithm,particle swarm optimization algorithm and improved them.The new algorithm is applied to image segmentation.The main research and works as follows:(1)Before making an image segmentation,combined with fractional order amplitude frequency characteristic curve and the fractional order calculus in the advantages of image denoising and enhancement.Aiming at the disadvantage that the fixed order of fractional order is not suitable for different feature images and needs to be manually selected,an adaptive image denoising and enhancement algorithm of fractional order calculus is proposed.According to the texture of the pixels in the image,the noise intensity should be treated with different fractional order,combined with the characteristics that fractional order can achieve good effect in a certain range,the fractional order adaptive adjustment formula with different gradient values is proposed.To ensure that the image where the gradient is bigger,the noise corresponds to the smaller negative order,while the edge corresponds to the larger positive order;the image where the gradient is smaller,the noise corresponds to the larger negative order,while the texture corresponds to the smaller positive order.At the same time it realizes the adaptive fractional order and improves image preprocessing effect.(2)Based on in-depth analysis of the traditional Otsu algorithm,the fractional calculus algorithm and the particle swarm optimization algorithm,three algorithms will be combined and improved,the Otsu image thresholding segmentation(ImFpsoOtsu)algorithm based on fractional order particle swarm optimization is proposed.First,this algorithm used the grayscale-gradient two-dimensional histogram,and inter-class variance of Otsu algorithm is defined as the fitness function.Second,by introducing the particle evolution factor and using the particle state information to adjust the fractional a,and the speed and the position are update by setting speed to zero.Finally,an improved particle swarm optimization algorithm combined with the traditional particle updating formula and using the particle symmetry distribution obtains the optimal threshold,and the target is segmented from the image.The fractional order particle swarm Otsu algorithm finally achieved the effective image segmentation.It solves the problem that the traditional particle swarm optimization algorithm falls into local optimum and improves the convergence speed.The experimental results show that the adaptive fractional order image preprocessing algorithm is superior to the traditional algorithm in terms of subjective vision and objective signal-to-noise ratio and entropy value,that is,image denoising,and the enhancement effect is better.Based on the Otsu algorithm of fractional particle swarm optimization,the convergence degree of the visual effect and the fitness curve shows that the proposed algorithm has faster convergence speed and higher accuracy.