改进动态阈值去噪在远程图像小目标检测中的应用研究  

Application of improved dynamic threshold denoising in remote image small-target detection

作　　者：刘冰[1] 

机构地区：[1]达州职业技术学院,四川达州635001

出　　处：《现代电子技术》2017年第2期81-84,共4页Modern Electronics Technique

基　　金：国家自然科学基金(51972712)

摘　　要：远程图像小目标在视频帧图像中占据的比例较低,无法有效划分目标和背景,目标检测精度低。因此,将改进动态阈值去噪法应用于远程图像小目标检测中,给出了远程图像小目标检测系统硬件总体结构,通过图像采集模块对图像传感器进行驱动,将采集到的远程图像数据保存至SDRAM模块中,并将该图像数据传输到显示器上显示。软件设计中,给出将改进动态阈值去噪方法应用于远程图像小目标检测的详细过程。实验结果表明,所提方法在检测远程图像小目标过程中具有可行性,可准确分割出小目标,满足检测实时性的需求,并且具有较强的适应能力。Since the remote image small target occupys a relatively low percentage in the video frame image,can＇t divide the target and background effectively,and has low target detection precision,the improved dynamic threshold denoising method is applied to the remote image small- target detection. The hardware overall structure of the remote image small- target detection system is given. The image sensor is driven by the image acquisition module to save the collected remote image data in SDRAM module,and transmit the image data to VGA for display. In software design,the detail process of applying the improved dynamic threshold denoising method to the remote image small-target detection is given. The experimental results show that the method is feasible in the process of remote image small-target detection,can segment the small target accurately,meet the needs of detection real-time performance,and has strong adaptive capability.

关 键 词：动态阈值 图像去噪 远程图像 小目标检测 

分 类 号：TN911.73[电子电信—通信与信息系统] TN219[电子电信—信息与通信工程]