摘要对红外目标进行检测是依靠对目标的成功辨认和跟踪,对红外目标检测算法的研究成为了近年研究的热门课题,成功的研究能有效提高红外预警系统、改善红外制导系统性能,因此对其进行深入研究有重要的理论意义和实用价值。研究出一个更加完善,更加优异的图像分割算法对红外目标检测是具有重大意义的。
本课题主要是在MATLAB语言环境下,编程实现常用的红外图像目标检测算法,例如Otsu算法,fcm算法,最佳阈值算法等不同算法在处理相同的单帧帧图像时生成的处理结果进行收集,除了上述有关阈值方面的,本文还提到了边缘检测算子的基本应用,利用各式各样的算法进行仿真,然后通过对主观的观察以及客观数据的讨论,并对其对比分析,探索硬件能够实时实现,判断出不同的算法的优缺点和适用范畴。64923
毕业论文关键词 红外图像 阈值分割 目标检测
毕业设计说明书(论文)外文摘要
Title the research of algorithm for infrared image segmentation
Abstract Infrared target detection and tracking of target recognition is based on infrared target detection algorithm research has been a hot research topic in recent years, which is to improve the infrared alarm system, infrared guidance system performance is the key, so its in-depth study has important theoretical significance and practical value. Come up with a better and more superior image segmentation algorithm for infrared target detection is of great significance.
The main subject is in the MATLAB language environment, the programming of commonly used infrared image target detection algorithms such as Otsu algorithm, fcm algorithm, the optimal threshold algorithm of different algorithms in dealing with the same single-frame image frame generated when processing the results of the collection, in addition to the above aspects of the threshold, the paper also mentions edge detection operators the basic application, the use of a variety of algorithms for the simulation, and then through subjective observation and discussion of objective data, and its comparative analysis, exploration hardware can real-time implementation, determine the advantages and disadvantages of different algorithms and application areas.
Keywords Image segmentation Threshold segmentation Target Detection
目次
1 引言 1
1.1图像与数字图像处理1
1.2 研究图像分割处理的意义 1
1.4本课题的研究工作2
2 图像分割算法与大致分类 3
2.1基于阈值的图像分割 3
2.2区域生长和分裂合并4
2.3 利用边缘检测算子5
3 基于阈值的图像分割算法概述 12
3.1 最大类间差算法13
3.2 模糊C均值聚类算法15
3.3二维熵分割算法16
3.4最佳阈值分割法19
3.5最大熵图像分割法20
3.6分水岭图像分割算法22
结论 24
致谢 26
参考文献27
1 引言
据研究,在人类能接收到的全部信息中,8成的信息是通过眼睛也就是视觉得到的,与读,听的辅助接受信息的方式相比,眼睛观察能接受信息量更加旁大,更加明显,更加客观,所以也就是更高效和更广泛的适应性。当我们手握一个充满了内容的图像时,对之进行合理的分割和处理,可以跟容易识别,分析,最后达到对信息的完全理解。