Remote sensing imageryaccuracy analysis method and apparatus

Abstract

A method of enhancing a resolution of an image by fusing images includes applying a principal component analysis to a multispectral image to obtain a plurality of principal components, and replacing a first component in the plurality of principal components by a panchromatic image. The method further includes resampling remaining principal components to a resolution of the panchromatic image, and applying an inverse principal analysis to the panchromatic image and the remaining principal components to obtain a fused image of the panchromatic image and the multispectral image.

1 . A method of evaluating the effects of image manipulation, such as sharpening and compressing an original image, by comparing the original image to the manipulated image results through use of a collection of quality metrics, the method comprising:
 applying a principal component analysis to a multispectral image to obtain a plurality of principal components;   replacing a first component in the plurality of principal components by a panchromatic image;   resampling remaining principal components to a resolution of the panchromatic image; and   applying an inverse principal analysis to the panchromatic image and the remaining principal components to obtain a fused image of the panchromatic image and the multispectral image.   
     
     
         2 . The method of  claim 1 , further comprising computing quality metrics on the fused image. 
     
     
         3 . The method of  claim 2 , wherein the computing of the quality metrics comprises computing a mean square error value between a band of the fused image and a band of the multispectral image. 
     
     
         4 . The method of  claim 2 , wherein the computing of the quality metrics comprises computing a root mean square value between a band of the fused image and a band of the multispectral image. 
     
     
         5 . The method of  claim 2 , wherein the computing of the quality metrics comprises computing a correlation between a first band in the multispectral image and a second band in the multispectral image and between a first band in the fused image and a second band in the fused image. 
     
     
         6 . The method of  claim 2 , wherein the computing of the quality metrics comprises computing a correlation between a band in the multispectral image and the panchromatic image and between a band in the fused image and the panchromatic image. 
     
     
         7 . The method of  claim 2 , wherein the computing of the quality metrics comprises computing a relative shift mean for each band of the fused image. 
     
     
         8 . The method of  claim 2 , wherein the computing of the quality metrics comprises computing histograms of bands in the multispectral image and computing histograms of bands in the fused image. 
     
     
         9 . The method of  claim 8 , further comprising comparing between a histogram of a band in the multispectral image and a histogram of a band in the fused image. 
     
     
         10 . The method of  claim 2 , wherein the computing of the quality metrics comprises computing an image noise index for each band of the fused image. 
     
     
         11 . The method of  claim 10 , wherein a negative value of the image noise index for a band corresponds to a degradation of spectral information for the band. 
     
     
         12 . The method of  claim 2 , wherein the computing of the quality metrics comprises computing a normalized difference vegetation index (NDVI) for the fused image and computing a normalized difference vegetation index for the multispectral image and correlating the normalized difference vegetation index for the fused image and the normalized difference vegetation index for the multispectral image. 
     
     
         13 . A method of evaluating the effects of image manipulation, such as sharpening and compressing an original image, by comparing the original image to the manipulated image results through use of a collection of quality metrics, the method comprising:
 applying a wavelet-based pansharpening to a plurality of bands in a multispectral image and a panchromatic image to obtain a pansharpened image; and   computing quality metrics on the pansharpened image.   
     
     
         14 . The method of  claim 13 , wherein the applying of the wavelet-based pansharpening comprises using a bi-orthogonal mother wavelet. 
     
     
         15 . The method of  claim 13 , further comprising applying filtering on the pansharpened image to remove noise. 
     
     
         16 . The method of  claim 15 , wherein applying the filtering comprises applying a Wiener filter on the pansharpened image. 
     
     
         17 . The method of  claim 13 , wherein the computing of the quality metrics comprises computing a root mean square value for each band of the pansharpened image. 
     
     
         18 . The method of  claim 13 , wherein the computing of the quality metrics comprises computing a correlation between a first band in the multispectral image and a second band in the multispectral image and between a first band in the fused image and a second band in the pansharpened image. 
     
     
         19 . The method of  claim 13 , wherein the computing of the quality metrics comprises computing a correlation between a band in the multispectral image and the panchromatic image and between a band in the fused image and the panchromatic image. 
     
     
         20 . The method of  claim 13 , wherein the computing of the quality metrics comprises computing a relative shift mean for each band of the pansharpened image. 
     
     
         21 . A method of evaluating the effects of image manipulation, such as sharpening and compressing an original image, by comparing the original image to the manipulated image results through use of a collection of quality metrics, the method comprising:
 preprocessing an image;   applying a discrete wavelet transform on the preprocessed image to decompose the preprocessed image into a plurality of sub-bands;   applying a quantization to each sub-band in the plurality of sub-bands;   partitioning the plurality of sub-bands into a plurality of code-blocks;   encoding each code-block in the plurality of code-blocks independently to obtain a code-blocks stream;   applying a rate control process to the code-blocks stream to obtain a bit-stream; and   organizing the bit-stream to obtain a compressed image.   
     
     
         22 . The method of  claim 21 , further comprising:
 transforming the compressed image using embedded block decoding to obtain embedded decoded block data;   re-composing the embedded decoded block data using an inverse discrete wavelet decomposition process;   performing a dequantization by assigning a single quantum value to a range of values to obtain a dequantized data; and   performing a decoding process on the dequantized data to substantially reconstruct the image.   
     
     
         23 . The method of  claim 22 , wherein the image has a tagged image file format (TIFF). 
     
     
         24 . The method of  claim 22 , wherein the image has a GeoTIFF format. 
     
     
         25 . The method of  claim 22 , wherein the applying of the discrete wavelet transform on the preprocessed image comprises decomposing each preprocessed image tile in a plurality of preprocessed image tiles into a high and low sub-bands of the preprocessed image tile with a low-pass filter and a high-pass filter. 
     
     
         26 . The method of  claim 22 , wherein the applying of the quantization to each sub-band in the plurality of sub-bands comprises assigning a range of values to a single quantum value in each sub-band. 
     
     
         27 . The method of  claim 22 , wherein the partitioning of the plurality of sub-bands into the plurality of code-blocks comprises partitioning the plurality of sub-bands into the plurality of code-blocks such that the code-blocks from each sub-band have substantially a same size. 
     
     
         28 . A method of evaluating the effects of image manipulation, such as sharpening and compressing an original image, by comparing the original image to the manipulated image results through use of a collection of quality metrics, the method comprising:
 inputting a GeoTIFF image file;   extracting a GeoTIFF header that contains references to geographic metadata;   creating a degenerated GeoTIFF image using the extracted geographic metadata;   performing a geographic markup language (GML) conversion;   inserting the degenerated GeoTIFF image into a universally unique identifier (UUID) box of the JP2 file;   inserting the geographic markup language into an extandible markup language (XML) box of the JP2 file; and   compressing the JP2 file using JP2000 image compression to obtain a GeoJPEG2000 image file.   
     
     
         29 . The method of  claim 28 , wherein the compressing using the JP2000 image compression comprises compressing with LuraWave.jp2 image compression code, JP2 Java/JNI-SDK or GeoTIFF-JAI. 
     
     
         30 . The method of  claim 28 , wherein the compressing using the JP2000 image compression comprises using a compression code developed using java. 
     
     
         31 . The method of  claim 28 , further comprising:
 decompressing the GeoJP2000 image file to obtain a decompressed TIFF image file; and   computing quality metrics to compare the GeoTIFF image and the decompressed TIFF image file.   
     
     
         32 . The method of  claim 31 , wherein the compressing of the JP2 file comprises compressing the JP2 file at a plurality of compression ratios. 
     
     
         33 . The method of  claim 32 , wherein the computing of the quality metrics comprises computing a mean square error value for each band of the decompressed image file at each compression ratio in the plurality of compression ratios. 
     
     
         34 . The method of  claim 33 , wherein as the compression ratio increases the mean square error value increases for each band. 
     
     
         35 . The method of  claim 32 , wherein the computing of the quality metrics comprises computing a root mean square value for each band of the decompressed image file at each compression ratio in the plurality of compression ratios. 
     
     
         36 . The method of  claim 35 , wherein as the compression ratio increases the root mean square error value increases for each band. 
     
     
         37 . The method of  claim 32 , wherein the computing of the quality metrics comprises computing a peak signal to noise ratio (PSNR) for each band of the decompressed image file at each compression ratio in the plurality of compression ratios. 
     
     
         38 . The method of  claim 37 , wherein as the compression ratio increases the peak signal to noise ratio decreases for each band. 
     
     
         39 . The method of  claim 32 , wherein the computing of the quality metrics comprises computing a correlation between a first band and a second band in the decompressed image file at each compression ratio in the plurality of compression ratios. 
     
     
         40 . The method of  claim 39 , wherein as the compression ratio increases the correlation between the first and second bands remains substantially constant.