Research On High Precision Measurement Of Speed Height Ratio For Aerial Camera

The precision of image motion compensation is the key factor impacting the imagequality. High precision measurement of speed height ratio (SHR) is the foundation forhigh precision image motion compensation. SHR measurement is converted to imagemotion measurement in this paper, aerial camera image motion measurement is studiedwhich is based on spatial filtering velocity measurement (SFV). Filter characteristics ofspatial filtering effect, filter parameter selection, two dimension velocity measurement,SFV signal process and related techniques are further studied.The validity of SFV for aerial camera is proposed comparing with optical velocitymeasurement. The principal and characteristics of SFV are introduced and spatialfiltering effect of image light intensity distribution is presented mathematically. Theeffects of transmittance function on characteristics of narrow band filter are discussed.The power spectrum presentations of typical transmittance function for different shapeof filter window are presented, and the characteristics of spectrum bandwidth andcentral frequency of the power spectrum are studied. The key parameters affectingidentity of spatial filter are analyzed by studying power spectrum function of spatialfilter. The applicability of various spatial filters, such as transmittance grating, prismgrating, optical-electoral diode array and liquid crystal component array, for imagemotion compensation is compared. The pedestal component elimination usingdifference demodulation is discussed. Direction discrimination of image motion andtwo dimension velocity measurements are proposed. SFV signal process and relatedsignal analysis techniques are further discussed. Typical signal analysis techniques,such as frequency spectrum analysis, frequency tracking and counting method, areintroduced.A novel method for image motion measure is proposed using spatial filteringeffect of line CCD for aerial camera. The narrow band filtering property of spatial splitis simulated through interval sampling line CCD image. An optical noncontact measurement of aerial camera image motion velocity is realized, and the SHRmeasurement is accordingly achieved. This method can flexibly adjust frequencycharacteristic of spatial filter according to frequency distribution of targets on ground,so it is applicable to high precision measurement of SHR for aerial camera. Severaltypical image motion velocities ranging from5mm/s to53.2mm/s are experimentallymeasured. The experimental results indicate that the maximum image motion errorresulting from measurement error is2μm when CCD pixel pitch is8μm and time ofexposure is10ms. Pixel binning of line CCD is utilized in the experiment, whichrealizes dynamic fast sampling and enhances image SNR. Direction identificationmethod for two-dimension velocity based on blur image is introduced and experimentis carried out to validate the method.