二维最小误差分割在红外图像中的快速实现

发布时间：2018-01-13 04:35

本文关键词：二维最小误差分割在红外图像中的快速实现　出处：《红外技术》2016年12期 　论文类型：期刊论文

【摘要】：最小误差分割算法的图像分割性能优异,但一维的最小误差分割算法容易受到噪声的干扰。利用图像的二维直方图,二维最小误差分割算法不仅能够利用图像的灰度信息,同时利用了相邻像素之间的邻域信息,取得更加理想的分割效果。但在实际使用的过程中,二维最小误差算法采用穷尽搜索的算法运算时间长,二维直线型最小误差分割算法无法反映全局最优解,降维形式的最小误差算法复杂度高。本文将结合粒子群优化算法(PSO)将二维最小误差分割算法应用在红外图像上,大大提升了算法的求解速度,能够在实现更低对比度的红外图像分割的同时满足工程中实时检测的要求。
[Abstract]:The minimum error segmentation algorithm has excellent image segmentation performance, but the one-dimensional minimum error segmentation algorithm is easily disturbed by noise. The two-dimensional histogram of the image is used. Two-dimensional minimum error segmentation algorithm can not only make use of the gray level information of the image, but also use the neighborhood information between adjacent pixels to achieve a more ideal segmentation effect, but in the process of practical use. Two-dimensional minimum error algorithm uses exhaustive search algorithm, and two-dimensional linear minimum error segmentation algorithm can not reflect the global optimal solution. In this paper, the two-dimensional minimum error segmentation algorithm is applied to infrared images, which greatly improves the speed of solving the algorithm, because of the high complexity of the minimum error algorithm in dimensionality reduction form, which combines the particle swarm optimization (PSO) algorithm with the particle swarm optimization (PSO) algorithm. It can achieve lower contrast infrared image segmentation and meet the requirements of real-time detection in engineering.
【作者单位】：中国科学院红外探测与成像技术重点实验室;中国科学院大学;
【分类号】：TP391.41
【正文快照】： 0引言图像分割旨在将图像分成若干特定的、具有独特性质的区域并提出感兴趣目标的技术和过程。它是图像处理和计算机视觉领域的基础部分,在目标检测、目标识别、特征提取、图像分类等领域都得到了广泛的应用,具有重要的研究价值。现有的图像分割主要分为以下几类[1]:基于熵的

【相似文献】