Noisy or corrupt images are characterized by poor contrast and ill-defined ridges and valleys. Such images need to be enhanced using suitable image enhancement techniques so that they can be used with other applications. Two improved image enhancement algorithms based on the switching median filtering are proposed in the present work. Images simulated with varying levels of Gaussian and salt-and-pepper noises are used for studying the efficacy of the proposed algorithms. A comparison with existing switching median filtering algorithm, in terms of visual quality of images, peak-signal-to-noise (PSNR) ratio and structural similarity index (SSIM), reveals the superiority of the proposed algorithms.

嘈杂或损坏的图像的特征是对比度差以及不清晰的山脊和山谷。需要使用适当的图像增强技术来增强此类图像，以便可以将其与其他应用程序一起使用。本文提出了两种基于切换中值滤波的改进图像增强算法。利用高斯噪声和椒盐噪声变化水平模拟的图像用于研究所提出算法的有效性。与现有的切换中值滤波算法的比较，在图像的视觉质量，峰值信噪比（PSNR）和结构相似性指数（SSIM）方面，揭示了所提出算法的优越性。