航空相机高精度速高比测量技术的研究
发布时间：2019-06-04 02:12
【摘要】：航空相机在空中飞行照相时，像移补偿精度是影响航空相机成像质量的关键因素，高精度像移补偿的前提是需要对速高比进行高精度的测量。本文将速高比测量转化为像移速度测量，从航空相机像移速度测量出发，，以空间滤波测速法（SFV）为基础，对空间滤波器的滤波特性、滤波器的参数选择、二维速度测量以及SFV信号处理等相关技术进行了深入研究。 通过对光学测速方法进行回顾，提出了利用空间滤波测速法对航空相机像移速度进行测量；介绍了空间滤波测速的原理和特性，并从数学角度对图像光强分布的空间滤波效应进行了介绍；讨论了透射函数对于窄带通滤波特性的影响，给出了不同形状的滤波器窗口对于典型透射函数的功率谱表达式，并对其功率谱中的频谱带宽及中心频率等特性进行了研究；通过研究空间滤波器的功率谱密度函数，对影响空间滤波特性的关键参数进行了分析；对比了透射光栅、棱镜光栅、光电二极管阵列以及液晶元件阵列等各类空间滤波器件进行像移速度测量的适用性；讨论了利用差分检波对基频成分的消除，提出了进行像移速度方向鉴别以及二维速度测量的方法；深入的讨论了处理SFV信号所涉及的信号分析技术，介绍了频谱分析，频率跟踪，计数方法以及相关技术等典型的信号分析技术。 提出了利用线阵CCD的空间滤波特性进行航空相机像移速度测量的新方法，利用线阵CCD的空间滤波特性，通过对线阵CCD推扫所产生的图像进行隔行采样，模拟了空间滤波器的窄带通频率特性，实现了对航空相机像移速度的光学非接触测量，从而实现了速高比的测量。此方法可根据地面目标的频率分布灵活调整空间滤波器的频率特性，能够适用于航空相机速高比的精确测量。实验对5mm/s~53.2mm/s范围内几个典型像移速度进行了测量，结果表明，CCD像元尺寸为8μm，相机曝光时间为10ms时，测量误差引起的像移量误差最大为2μm，能够满足航空相机像移补偿的精度要求；实验中采用了线阵CCD的像素binning功能，大幅度提高了数据的读出速率，实现了动态累加的快速采样，同时提高图像的信噪比；提出了利用单个线阵CCD进行一维运动方向鉴别的方法，简化了光学系统；介绍了基于模糊图像的二维速度方向鉴别方法，并进行了试验验证。 本文成功的实现了航空相机像移速度的高精度测量，从而实现了速高比的高精度测量，试验结果表明，文中提出的测量方法其精度完全满足航空相机像移补偿的精度要求。
[Abstract]:When aerial camera is photographed in the air, the accuracy of image shift compensation is the key factor affecting the imaging quality of aerial camera, and the premise of high precision image shift compensation is the need to measure the ratio of speed to height with high precision. In this paper, the measurement of speed to height ratio is transformed into image shift velocity measurement. Based on the spatial filtering velocimeter method (SFV), the filtering characteristics of the spatial filter and the parameter selection of the filter are based on the image shift velocity measurement of the aerial camera. The related technologies such as two-dimensional velocity measurement and SFV signal processing are deeply studied. Based on the review of the optical velocimeter method, the spatial filtering velocimeter method is proposed to measure the image shift velocity of the aerial camera. The principle and characteristics of spatial filtering velocity measurement are introduced, and the spatial filtering effect of image light intensity distribution is introduced from the mathematical point of view. The influence of transmission function on narrow band pass filtering characteristics is discussed. The power spectrum expressions of filter windows with different shapes for typical transmission functions are given, and the characteristics of spectrum bandwidth and center frequency in the power spectrum are studied. By studying the power spectral density function of the spatial filter, the key parameters affecting the spatial filtering characteristics are analyzed. The applicability of various spatial filter devices, such as transmission grating, prism grating, photodiode array and liquid crystal element array, for image shift velocity measurement is compared. The elimination of fundamental frequency components by differential detection is discussed, and the methods of image shift velocity direction identification and two-dimensional velocity measurement are proposed. The signal analysis technology involved in processing SFV signal is discussed in detail, and the typical signal analysis techniques, such as spectrum analysis, frequency tracking, counting method and related technology, are introduced. A new method for measuring the image shift velocity of aerial camera by using the spatial filtering characteristics of linear CCD is proposed. By using the spatial filtering characteristics of linear CCD, the images generated by linear CCD push sweep are sampled interlaced. The narrowband frequency characteristics of the spatial filter are simulated, and the optical non-contact measurement of the image shift velocity of the aerial camera is realized, thus the measurement of the speed to height ratio is realized. This method can flexibly adjust the frequency characteristics of the spatial filter according to the frequency distribution of the ground target, and can be applied to the accurate measurement of the speed to height ratio of the aerial camera. Several typical image displacement speeds in the range of 5mm/s~53.2mm/s are measured experimentally. the results show that when the size of CCD pixel is 8 渭 m and the camera exposure time is 10ms, the maximum image shift error caused by the measurement error is 2 渭 m. It can meet the accuracy requirements of image shift compensation of aerial camera. The pixel binning function of linear CCD is used in the experiment, which greatly improves the readout rate of the data, realizes the fast sampling of dynamic accumulation, and improves the signal-to-noise ratio of the image at the same time. A method of one-dimensional motion direction identification based on single linear CCD is proposed, which simplifies the optical system, and the two-dimensional velocity direction identification method based on fuzzy image is introduced and verified by experiments. In this paper, the high precision measurement of image shift speed of aerial camera is successfully realized, and the high precision measurement of speed to height ratio is realized. The experimental results show that the accuracy of the measurement method proposed in this paper fully meets the accuracy requirements of image shift compensation of aerial camera.
【学位授予单位】：中国科学院研究生院（长春光学精密机械与物理研究所）
【学位级别】：博士
【学位授予年份】：2013
【分类号】：P231
[Abstract]:When aerial camera is photographed in the air, the accuracy of image shift compensation is the key factor affecting the imaging quality of aerial camera, and the premise of high precision image shift compensation is the need to measure the ratio of speed to height with high precision. In this paper, the measurement of speed to height ratio is transformed into image shift velocity measurement. Based on the spatial filtering velocimeter method (SFV), the filtering characteristics of the spatial filter and the parameter selection of the filter are based on the image shift velocity measurement of the aerial camera. The related technologies such as two-dimensional velocity measurement and SFV signal processing are deeply studied. Based on the review of the optical velocimeter method, the spatial filtering velocimeter method is proposed to measure the image shift velocity of the aerial camera. The principle and characteristics of spatial filtering velocity measurement are introduced, and the spatial filtering effect of image light intensity distribution is introduced from the mathematical point of view. The influence of transmission function on narrow band pass filtering characteristics is discussed. The power spectrum expressions of filter windows with different shapes for typical transmission functions are given, and the characteristics of spectrum bandwidth and center frequency in the power spectrum are studied. By studying the power spectral density function of the spatial filter, the key parameters affecting the spatial filtering characteristics are analyzed. The applicability of various spatial filter devices, such as transmission grating, prism grating, photodiode array and liquid crystal element array, for image shift velocity measurement is compared. The elimination of fundamental frequency components by differential detection is discussed, and the methods of image shift velocity direction identification and two-dimensional velocity measurement are proposed. The signal analysis technology involved in processing SFV signal is discussed in detail, and the typical signal analysis techniques, such as spectrum analysis, frequency tracking, counting method and related technology, are introduced. A new method for measuring the image shift velocity of aerial camera by using the spatial filtering characteristics of linear CCD is proposed. By using the spatial filtering characteristics of linear CCD, the images generated by linear CCD push sweep are sampled interlaced. The narrowband frequency characteristics of the spatial filter are simulated, and the optical non-contact measurement of the image shift velocity of the aerial camera is realized, thus the measurement of the speed to height ratio is realized. This method can flexibly adjust the frequency characteristics of the spatial filter according to the frequency distribution of the ground target, and can be applied to the accurate measurement of the speed to height ratio of the aerial camera. Several typical image displacement speeds in the range of 5mm/s~53.2mm/s are measured experimentally. the results show that when the size of CCD pixel is 8 渭 m and the camera exposure time is 10ms, the maximum image shift error caused by the measurement error is 2 渭 m. It can meet the accuracy requirements of image shift compensation of aerial camera. The pixel binning function of linear CCD is used in the experiment, which greatly improves the readout rate of the data, realizes the fast sampling of dynamic accumulation, and improves the signal-to-noise ratio of the image at the same time. A method of one-dimensional motion direction identification based on single linear CCD is proposed, which simplifies the optical system, and the two-dimensional velocity direction identification method based on fuzzy image is introduced and verified by experiments. In this paper, the high precision measurement of image shift speed of aerial camera is successfully realized, and the high precision measurement of speed to height ratio is realized. The experimental results show that the accuracy of the measurement method proposed in this paper fully meets the accuracy requirements of image shift compensation of aerial camera.
【学位授予单位】：中国科学院研究生院（长春光学精密机械与物理研究所）
【学位级别】：博士
【学位授予年份】：2013
【分类号】：P231
【参考文献】
相关期刊论文 前10条
1 卢万欣;;采用CCD线阵做探测器的数字相关测速系统[J];长春光学精密机械学院学报;1992年02期
2 张林;吴晓琴;袁
本文编号：2492398
本文链接：https://www.wllwen.com/kejilunwen/dizhicehuilunwen/2492398.html
