Sky vision 技术研究毕业论文
2022-04-12 19:56:43
论文总字数:17522字
摘 要
近些年来随着电子科技技术的不断发展,我们的摄影途径也由以往的专业相机设备向手机数码相机转变,人们对于摄影的兴趣也愈发浓厚。但是由于单个摄像机所拍到的画面视觉范围十分有限,难以呈现出一个宽场景所有的图像信息,为了能够得到全景图像,各种研究成果能够通过图像拼接技术来获得最终的全景场景,目前也有了成熟的模型与计算方法。
不过在实际的图像拼接操作中,传统的图像拼接技术并不能很好地实现图像之间的有效融合,往往会出现图像的噪点或者图像的清晰度会降低,并且也有着巨大的计算量,一定程度上增加了我们工作的难度与成本。本文提出的sky vision是将一个大的宽场景分化成多个独立画面,通过对其每个画面的像素点的RGB值与坐标的计算,再通过坐标系之间的转换,得出每个像素点与最终场景中的坐标位置对应关系。最终通过每个像素点RGB值与位置的合成,可以得到一个完整的宽大场景的图像。
Sky vision在技术方面有着广泛的应用。混合场景,可以实现跟踪对象。例如,它可以解决机器人跟踪清洁和其他任务。大场景,它可以迅速抓住全景的所有信息。例如,在开车的过程中,技术可以实现漂移修正车辆偏离主要道路。
关键词:宽场景;sky vision;像素点;RGB值
Sky Vision Technology
Abstract
Following the development of electronic technique in recent years, the ways of photography changed from professional photographic equipment to smart phone and digital camera. Thus, people are more and more interested in photography. But due to the limited visual range of signal camera, its difficult to present the all image information of a wide space. For getting the panoramic picture, various research results can be used to get this target by image mosaic technology, and there are the mature model and calculating methods now.
But during the actual operation, the image mosaic technology can’t fuse these images very well. There always have the noise or low definition, and a huge calculated amount. It increases the difficult and cost of our work partly. The sky vision technology mentioned in this thesis will divide a wide space into several independent images, through calculating the RGB value and coordinate of the pixel points of every image, then getting corresponding relations between each pixel point and coordinate position in the final scene by coordinate transformation. Finally we can get a whole image of the wide space through compounding the RGB values and positions.
Sky vision technology has a wide use nowadays. Combined scenes can be used to track objects. For example, it can be used to finish the tracking cleaning and other tasks of robots. The wide scene can catch the all information of full view quickly. For example, in driving process, this technology can do the drift correction.
Keywords: large wide scene;sky vision; pixel points; RGB value
目 录
摘 要 I
Abstract II
第一章 绪论 1
1.1课题研究的背景和意义 1
1.2图像拼接技术应用 1
1.2.1 虚拟现实技术应用 2
1.2.2 遥感技术的应用 2
1.2.3全景监控系统应用 3
1.3图像拼接技术的研究现状 4
1.4本论文工作梳理 5
第二章 基于局部特征提取的图像拼接方法 7
2.1 图像的预处理 7
2.1.1图像灰度法 7
2.1.2图像二值化处理 8
2.2 一般局部特征提取算法 9
2.2.1 Moravec角点检测 9
2.2.2 Harris角点检测 9
2.2.3 LoG边缘检测算法 10
2.3 SIFT特征检测算法 10
2.4实验结果与分析 11
第三章sky vision原理与算法 13
3.1平行计算和透视投影 13
3.2 图像离散化处理 15
3.3 sky vision的算法 16
3.4 sky vision合成结果与分析 19
3.5本章工作小结 20
第四章 总结与未来工作展望 21
参考文献 23
致 谢 25
请支付后下载全文,论文总字数:17522字