Evaluation of an Airborne Optical Remote Sensing System and Operational Pre-Processing Methods

다국어 초록 (Multilingual Abstract)

Fo remote sensing (RS)to be a truly useful technique in environmental science,the remotely sensed data need to be of known quality and reliability. The various step known collectively as 'pre-processing'are vitally important as erro and uncertainty at this stage has the potentialcause large errors in the final data product. Two important aspects of pre- processing were investigated in this research First the radiometic conversion by which acquired signal are converted into physically meaningful values and which allows the comparison on of dataset from different sensors or from different dates with the same sensor was studied.Second, the process by which the influence of the atmosphere is removed from the remotely sensed signal was investigated, focusing upon practical method to correct data collected by CASI-2, an imaging spectromete produced by Itres Research.
The radiomtric (and other) characteristics of any sensor or system are normally obtained by laboratory calibration. Several experiments were conducted to investigate and enhance knowledge of the performance of the CASI-2. The results suggested that wavelength calibration could reveal systematic optical distortion due to the'optical smile' effect, and that the uncertainty of the wavelength calibration could be reduced if this was taken into accocount. Another possible source of error in the sensor calibration was identified and traced to spatially non-uniform radiance standard, improvem of which could greatly reduce systematic error across the image. In addition to investigation of the conventional sensor calibration, several studies were conducted to acquire new information. For example, the spectral response function of the CASI-2 was determined independently of the manufacture for the first time, using an innovative iterative procedure.
In addition to the research conducted on assessing the laboratory calibration procedure of the CASI, and investigatingits performance, a series of laboratory investigations were undertaken characterise the CASI IncidentLight Sensor (IL ILS). Genenerally, the performance of the ILS, such its angular response and radiometric linearity were acceptable for the purpose for which it was designed. However, signals in the short wavelengt region seemed to sufferr from relativelylow signal-to-noise ratio.
In the interest of operational aspects of airborne multispectral RS, the contribution of atmospheric variation to remotely sensed was reviewed, focusing,in particular,upon numerical models developed to characterise the sky radiance distribution. Atmospheric effects on remotely sensed data were reviewed and the effects of atmospheric variability studied terms of how this influences the remotely sensed signal. Temporal variation in atmospheric clarity (and by extension spatial variations typical of RS image data),were shown to cause errors which also affected multispectral ratio-based analysis.Two novel practical methods of atmospheric correction were developed following series of practical experiment and theoretical studies.
Finally,two novel practical methods of correcting remotely sensed data for the effect of the atmosphere were presented.The first was an image-based method which can be used to achieve a qualitative improvement in image quality, reduction in scene noise where this is due to shadowing and an improvement in accuracy of multispectral ratios. The second described an innovative way to use the data on downwelling irradiance measured by a roof- mounted sensor to correct for irradiance radiation affecting RS data. The method also has the potential to provide data on the sky irradiance distribution at thetime of sensing.

목차 (Table of Contents)

• DECLARATION OF AUTHORSHIP = ii
• ABSTRACT = iii
• ACKNOWLEDGEMENTS = v
• TABLE OF CONTENTS = vi
• LIST OF FIGURES = x
• LIST OF TABLES = xxiii
• GLOSSARY = xxv
• Chapter 1 Introduction = 1
• 1.1. Remote Sensing for Environmental Science = 1
• 1.2. Environmental variables of interest = 3
• 1.3. From data to information = 5
• 1.4. Sources of uncertainty = 7
• 1.5. Pre-processing of remote sensing data = 10
• 1.6. Outline of the thesis = 14
• Chapter 2 Airborne multispectral remote sensing as an operational tool = 18
• 2.1. Operational aspects of remote sensing = 18
• 2.1.1. Spaceborne remotesensing = 18
• 2.1.2. Airborne sensing = 21
• 2.2. Multispectral remote sensing systems in airborne platforms = 26
• 2.2.1. Airborne simulators = 26
• 2.2.2. Hyperspectral sensors = 28
• 2.2.3. Multi-purpose and experimental sensors = 32
• 2.3. Summary = 33
• Chapter 3 Description of CASI system = 36
• 3.1. Operation of CASI-2 in the UK = 36
• 3.2. The Compact Airborne Spectrographic Imager (CASI-2): Overview = 40
• 3.2.1. The Instrument Control Unit = 42
• 3.2.2. The Sensor Head Unit = 43
• 3.2.3. The Incident Light Sensor = 46
• 3.3. Operation Modes = 47
• 3.3.1. Considerations of operation modes = 47
• 3.3.2. Spatial mode = 50
• 3.3.3. Original spectral mode = 51
• 3.3.4. Enhanced spectral mode = 52
• 3.3.5. Full-frame mode = 53
• 3.4. System Specification = 54
• 3.4.1. Signal related specifications = 56
• 3.4.2. Image related specifications age = 59
• 3.4.2.1. Spatial properties = 59
• 3.4.2.2. Spectral properties = 60
• 3.4.2.3. Radiometric properties = 62
• 3.5. Laboratory Calibration of CASI-2 System = 63
• 3.6. Summary = 67
• Chapter 4 Limitations of radiometric and spectral performance of the CASI-2 = 69
• 4.1. Introduction = 69
• 4.2. Spectral anisotropy across full spatial width = 71
• 4.2.1. Spectral data acquisition in full-frame mode ode = 72
• 4.2.2. Sub-pixel position of the spectrum line = 75
• 4.2.3. Spatial variation of wavelength calibration = 81
• 4.3. Case study: Sensitivity problem with wavelength calibration = 86
• 4.3.1. Raw CASI-2 image = 88
• 4.3.2. Differences of RSC between Itres and NERC27 = 89
• 4.3.3. Wavelength test between Itres and N27 = 90
• 4.4. Spectral response function of the CASI-2 = 93
• 4.4.1. The emission peaks in fine spectral resolution data = 94
• 4.4.2. Estimation of SRF of CASI-2 = 97
• 4.5. Non-uniformity problem on r radiometric calibration = 103
• 4.5.1. Merging two uniformity datasets = 104
• 4.5.2. Spatial non-uniformity due to integrating sphere = 108
• 4.6. Summary = 112
• Chapter 5 Investigation of the CASI-2 downwelling irradiance sensor = 114
• 5.1. Introduction = 114
• 5.2. Experimental design for the ILS investigations = 116
• 5.2.1. Linearity measurements = 119
• 5.2.2. Angular asurements = 120
• 5.2.2.1. Collimator = 120
• 5.2.2.2. Performance check = 121
• 5.2.2.3. Angular response of the ILS = 124
• 5.3. Linearity check = 127
• 5.4. Angular response of the ILS = 138
• 5.5. Summary = 141
• Chapter 6 Atmospheric and surface interactions in remotely sensed images = 144
• 6.1. Introduction = 144
• 6.2. Radiance source in optical remote sensing = 147
• 6.2.1. Interactions with the atmosphere = 148
• 6.2.1.1. Absorption = 150
• 6.2.1.2. Scattering = 151
• 6.2.2. Various paths of radiance = 151
• 6.2.3. Interactions with ground objects = 155
• 6.3. Inherent vs. apparent reflect reflectance of the surface = 158
• 6.3.1. Anisotropy of sky radiance distribution = 159
• 6.3.2. The adjacency effect = 165
• 6.3.3. Target BRDF = 166
• 6.4. Minimising of atmospheric effects = 168
• 6.4.1. Direct method from empirical measurements = 170
• 6.4.1.1. Ground based: Empirical line methods = 170
• 6.4.1.2. Multiple observations over thesame area = 172
• 6.4.1.3. Ancillary measurements sensor level = 175
• 6.4.2. Image-based methods = 177
• 6.4.2.1. Image Space Based = 178
• 6.4.2.2. Spectral domain based = 179
• 6.4.2.3. Feature Space Based = 181
• 6.4.3. Radiative Transfer Model-Based Methods = 186
• Chapter 7 Empirical measurements of short-term temporal variability of the atmosphere and development of a conceptual model = 191
• 7.1. Introduction = 191
• 7.2. Field data acquisition = 193
• 7.2.1. Instruments = 193
• 7.2.2. Field measurements with the ASD and MMR = 196
• 7.2.3. Spectral variations under sunlit and shadow boundaries (SSB) = 199
• 7.3. Slope and intercept of linear regression = 202
• 7.4. Linearity test = 205
• 7.5. Shadow-line model in 2D feature space = 210
• 7.5.1. Attributes of shadow-line with temporal variation of atmospheric conditions = 214
• 7.5.2. Nonlinearity of the shadow-line = 219
• 7.6. Dark Point Virtual Endmember (DPVE) = 221
• 7.6.1. A pair of shadow-lines = 222
• 7.6.2. More than two shadow-lines: The SVD = 224
• 7.6.3. The DPVE and the atmospheric effects = 226
• 7.7. Summary = 228
• Chapter 8 Image-based method: Sensing in shadows = 232
• 8.1. Introduction = 232
• 8.2. Data sets used = 236
• 8.2.1. Image Descriptions = 237
• 8.2.2. Shadows in the sample images = 238
• 8.2.2.1. Geometric shadow = 238
• 8.2.2.2. Ephemeral shadow = 240
• 8.3. Conventional shadow suppression methods = 241
• 8.3.1. RGB-HSI transformation methods = 242
• 8.3.2. Spectral Angular Mapper (SAM) = 246
• 8.4. Application of the DPVE on the Scene Shadows = 248
• 8.4.1. Attributes of the DPVE in image = 248
• 8.4.2. Shadows in 2-D feature space = 250
• 8.4.3. Shadows in n--dimensional space = 254
• 8.5. Summary = 262
• Chapter 9 Ancillary measurement based method: At-sensor downwelling irradiance = 265
• 9.1. Introduction = 265
• 9.2. Attitude correction of the ILS = 267
• 9.2.1. Attitude and navigation records: An example = 270
• 9.2.2. The Sun angle = 271
• 9.2.2.1. Flight navigation = 271
• 9.2.2.2. Computation of the solar angle = 271
• 9.2.3. Instantaneous ILS position = 272
• 9.2.3.1. The maximum slope (zenith angle) of the ILS = 273
• 9.2.3.2. Azimuthal angle of the ILS slope = 275
• 9.2.3.3. Absolute slope position with real data = 276
• 9.2.4. The ILS position vs. solar angle = 278
• 9.2.4.1. Solar zenith on an inclined surface = 279
• 9.2.4.2. Solar azimuth on an inclined surface = 280
• 9.3. Sky models for extreme sky conditions = 282
• 9.3.1. A single sky condition model = 282
• 9.3.2. Flexible sky condition model = 285
• 9.4. Sky models for all weather conditions = 288
• 9.4.1. Brunger and Hooper's SRDM = 288
• 9.4.2. Data Description = 290
• 9.4.3. Results = 292
• 9.4.4. ILS data over a calibration site = 296
• 9.4.5. Comparing ILS data in different flight directions = 298
• 9.5. Summary = 301
• Chapter 10 Conclusions = 305
• 10.1. Introduction = 305
• 10.2. Sensor calibration and radiometric correction = 307
• 10.3. Influence of atmospheric variations on images = 312
• 10.4. Concluding remarks and future studies = 317
• Appendix A Sensor system: Laboratory Calibration of the CASI-2 = 321
• References = 347