一种新的红外图像快速分割方法研究

研究红外图像目标分割快速优化问题。在高分辨率红外图像的分割中,红外图像存在数据量大、目标的边缘模糊和噪声较大等导致分辨率低和实时性差。为了快速准确分割,提出基于低尺度分割阈值预测的快速红外图像分割方法(LFIRS)。首先建立尺度阶数计算模型以确定保留原始红外图像目标基本信息所需的最小尺度,在分析多尺度过程对具有目标/背景强相关特性的红外图像进行分割阈值,建立多尺度红外图像分割阈值的相关模型(CLM),结合经典二维阈值法得到的最低两个尺度的阈值CLM模型参数,可以通过CLM模型与最小尺度的快速获取原始尺度的分割阈值,实现红外图像的快速分割。实验结果表明,改进方法提高了图像分辨和分割速度,且改善了分割效果。

New Method of Fast Infrared Image Segmentation

Through the study of the fast segmentation of high-resolution infrared image,the method of two-dimension threshold segmentation takes a long time to slove the original infrared image was found.To slove the ploblem,a method of fast segmentation of infrared Image based on low-scale thresholds(LFIRS) was proposed.First,the scale calculation model was established to obtain the minimum scale,and the scale preserved the orinigal infrared image information.The partition of segmentation threshold based on the strong correlation of infrared images was discussed.Then,a multi-scale infrared image segmentation threshold correlation model(CLM) was established.Finally,the original scale threshold was obtained quickly through the CLM and the small scale thresholds.Experiment results show that this method is better than the original method of segmentation in terms of speed and accuracy.