【作者】 赵芳；

【导师】 栾晓明；

【作者基本信息】 哈尔滨工程大学 ， 信号与信息处理， 2009， 硕士

【摘要】 水下成像技术在搜索、援救、探雷、科学成像、目标跟踪、导航控制和目标识别方面有着广泛的应用。本论文目的是针对水下球状物体成像图像进行处理、分析,估计图像曝光度与清晰度,以为实现自适应成像提供控制参数,使得在水下所成的图像能够自适应的调节清晰。根据水下图像的特点,本论文主要在图像预处理、图像目标的识别、自动曝光和自动调焦的算法等方面做了研究工作。本文首先对水下成像技术的概况进行介绍,重点分析水下成像特点。然后进行算法设计,对图像噪声的分析。对图像进行预处理,包括去除噪声、二值化处理、边缘处理。再从目标识别的角度,应用一种基于Pascal定理的椭圆检测方法,利用椭圆的几何性质,对随机选取的点进行两步筛选。筛选出有效椭圆边缘,然后再进行Hough变换,并给出结果。算法计算量小,计算时间短,占用的内存空间小。最后,以图像灰度直方图为基础,综合设计了一种曝光量函数计算方法。实验结果表明,该曝光量函数计算算法能够计算估计出图像的曝光效果,无需数据库支撑,复杂度低,易于实现。同时,依据各种图像梯度的分布情况,设计一种改进的图像清晰度评价函数,提出一种基于Prewitt评价函数的新的聚焦方法,并给出了实验分析和结果。曝光估计与清晰度评价是自适应成像的关键环节。更多还原

【Abstract】 Underwater imaging technology has a wide range of applications, which in the search, rescue, mine detection, scientific imaging, target tracking, navigation control and target identification. The purpose of the thesis is the Ray that underwater objects. In order to make the underwater image to adaptive control and make the underwater image to clear adaptive regulation, we processing and analysis the underwater imaging. Due to the characteristics of the underwater image, research work included image processing, image target distinguish, auto exposure algorithm and auto-focusing algorithm.First of all, a survey of underwater imaging technology development is reviewed, which emphasizes particularly on analysis of the problems existing in the underwater light. Then the preprocess of image is the base of following image processing, it mainly includes remove noise in image, binary image, image edge processing.And one application ellipse detection method is presented from the aspect of image target recognition, which based on theorem of Pascal, and utilized geometric properties of ellipse fully, select the random-dot on two step. After the edge of the ellipse is screened out efficaciously, Hough transformation are made, the method has the characteristics of less calculation, short computation time and taking up memory space. In the end, on the basis of the image gray histograms, the effect of the exposure images can estimate. Experimental results show that the algorithm can calculate the exposure of images effective, and need not database support, low complexity, and easy-to-implement. At the same time, it is found that the blurred image has large amount of pixels with low gradient. A kind of improved clarity-evaluation function is introduced. A new auto-focusing algorithm is proposed, which is based on Prewitt algorithm and it presented mathematical model and experimental analysis and result. Estimation of exposure and clarity-evaluation function is the key to adaptive control of image.更多还原