本文選題：多譜段 + 成像器件�。� 參考：《南京大學》2017年碩士論文

【摘要】：傳統(tǒng)的可見光相機在成像過程中只記錄了(R,G,B)三個顏色通道的分量,對于人眼來說這是足夠的,但是這遠不能精確表示光線在各個波長分量上的響應(yīng),而且容易受環(huán)境光的影響,安全監(jiān)控、醫(yī)療診斷等諸多領(lǐng)域都很受限制,非可見光波段成像器件以及光譜成像器件可以記錄(R,G,B)以外的譜段信息,可以彌補普通可見光相機的缺陷,但是目前的相機在成像過程中的投影變換使得所得照片失去了深度(z)信息,這使得利用單相機進行定位、診斷等工作也需要特殊的校準等工作才能進行;利用多光譜成像器件進行軍事安全、環(huán)境監(jiān)測、生物科學方面的工作也需要進行精準的校準工作。本文圍繞如何對各波段成像傳感器以及高光譜成像器進行準確校準展開研究,主要工作和創(chuàng)新點包括以下幾個方面:1.提出了一種對日盲紫外成像器件進行校準并進行定位的方法,與常用方法相比,能夠克服日盲紫外成像器件因分辨率以及色彩信息缺乏而導(dǎo)致的問題,能夠在參考點丟失的情況下進行補,使系統(tǒng)更為魯棒,實現(xiàn)了日盲紫外器件高精度校準與定位。2.提出了一種利用雙目紅外成像器件進行高精度定位反光醫(yī)學小球的方法,針對具體的應(yīng)用場景,能夠更準確地計算成像小球的空間坐標,使得定位更加準確,實現(xiàn)了紅外器件高精度全自動化校準與定位。3.提出了一種針對棱鏡-掩膜式多光譜采集相機的校準方式,與常用方法相比,能夠準確進行雙相機配準,提高光譜分辨率和和整個穩(wěn)定性,對于光譜數(shù)據(jù)的準確性提高有顯著的作用。
[Abstract]:Traditional visible-light cameras only record the components of the three color channels in the imaging process, which is sufficient for the human eye, but it is far from accurate enough to represent the response of the light to each wavelength component, and is vulnerable to the influence of ambient light. Security monitoring, medical diagnostics and many other areas are very restricted. Non-visible light band imaging devices and spectral imaging devices can record spectrum information other than the RPG-Bs, and can make up for the defects of ordinary visible light cameras. However, the current projection transformation of the camera in the imaging process results in the loss of the depth of the image information, which makes the use of a single camera for positioning, diagnosis and other work also needs special calibration work. The use of multispectral imaging devices for military security, environmental monitoring, and biological science also requires accurate calibration. This paper focuses on the accurate calibration of imaging sensors and hyperspectral imagers in each band. The main work and innovations include the following aspects: 1. A method for calibrating and locating solar blind ultraviolet imaging devices is proposed. Compared with common methods, it can overcome the problems caused by the lack of resolution and color information of solar blind ultraviolet imaging devices. It can make the system more robust when the reference point is lost, and realize the high-precision calibration and positioning of sunblind ultraviolet devices. In this paper, a method of high-precision positioning of reflective medical pellets with binocular infrared imaging devices is proposed. In view of the specific application scenarios, the spatial coordinates of imaging spheres can be calculated more accurately, and the positioning is more accurate. The high precision automatic calibration and positioning of infrared devices is realized. A calibration method for prism-mask multispectral acquisition camera is proposed. Compared with common methods, it can accurately register double cameras, improve spectral resolution and overall stability. It plays an important role in improving the accuracy of spectral data.
【學位授予單位】：南京大學
【學位級別】：碩士
【學位授予年份】：2017
【分類號】：TP212

【參考文獻】

相關(guān)期刊論文 前5條

1 胡超;葛云;陳穎;;人體關(guān)節(jié)活動度測量系統(tǒng)[J];中國醫(yī)學物理學雜志;2016年01期

2 周波;楊劍;王東平;;基于隨機采樣一致性算法的平面匹配方法[J];計算機應(yīng)用;2011年04期

3 夏仁波;劉偉軍;趙吉賓;徐金亭;;基于圓形標志點的全自動相機標定方法[J];儀器儀表學報;2009年02期

4 胡占義,吳福朝;基于主動視覺攝像機標定方法[J];計算機學報;2002年11期

5 李華,吳福朝,胡占義;一種新的線性攝像機自標定方法[J];計算機學報;2000年11期

，

本文編號：1929405