基于粒子群优化的BP神经网络图像复原算法研究
Image Resteoration by BP Neural Based on PSO

立足于粒子群算法与BP神经网络算法相结合的PSO-BP算法，在对其进行优化的基础上，将这一算法应用到图像复原的研究中。在PSO-BP优化算法模型中，一方面用BP算法将各个训练样本的误差进行反传，并用原始图片作为参考共同修正BP算法的权阈值；另一方面又通过正向粒子群算法及BP自身算法对复原图像进行优化。最后通过算法分析和实验数据验证PSO-BP优化算法的复原效果优于同类型算法。
Based on PSO-BP algorithm combining particle swarm algorithm with BP neural network algorithm, this paper applies this algorithm to image restoration based on optimization. In the PSO-BP optimization algorithm model, on the one hand, the error of each training sample of BP algorithm is reversed, and the original image is used as the reference to modify the weight threshold of BP algorithm. On the other hand, it is optimized by forward particle swarm algorithm and BP algorithm. Finally, through the algorithm analysis and experimental data, the recovery effect of PSO-BP optimization algorithm is better than that of the same type algorithm