基于轻量化YOLOX-S与多阈值分割的矿山遥感图像去噪算法

Mine Remote Sensing Image Denoising Algorithm Based on Lightweight YOLOX-S and Multi-threshold Eegmentation

矿山遥感图像普遍存在大量的噪点,给后续图像分析和处理带来了很大困难。提出了一种基于轻量化目标检测模型YOLOX-S和多阈值分割的矿山遥感图像去噪算法。首先使用YOLOX-S模型对矿山遥感图像进行目标检测,得到矿山目标的位置信息。然后针对矿山目标的特点,设计了一种多阈值分割方法消除图像中的噪声点。通过将图像分为若干个子区域,并对每个子区域采用不同的阈值进行二值化处理,最终将各子区域的二值化结果合并得到去噪后的图像。试验结果表明:该算法能够有效地去除矿山遥感图像中的噪声点,并且在保留目标特征的同时,大幅提升了图像质量。此外,由于采用了轻量化模型和多阈值分割算法,使得该算法具有较快的处理速度和较低的计算成本,适用于大规模图像数据的处理任务。

There are a lot of noise points in mine remote sensing images,which brings great difficulties to the analysis and processing of subsequent images.A denoising algorithm of mine remote sensing image based on lightweight object detection model YOLOX-S and multi-threshold segmentation is proposed.Firstly,YOLOX-S model is used to detect the mine remote sensing image,and the location information of the mine target is obtained.Then,according to the characteristics of the mine target,a multi-threshold segmentation method is designed to eliminate the noise in the image.The image is divided into several subregions,and each subregion is binarized with different thresholds.Finally,the binarized results of each subregion are combined to get the denoised image.The experimental results show that the algorithm can effectively remove the noise in the mine remote sensing image,and greatly improve the image quality while retaining the target features.In addition,due to the use of lightweight model and multi-threshold segmentation algorithm,the algorithm has a fast processing speed and low computing cost,which is suitable for large-scale image data processing tasks.