Digital Image Processing of Remote-Sensed Data
Digital Image Pre-Processing
Image pre-processing is a set of techniques employed to enhance the quality and extract relevant information from digital images before they are further analyzed and processed by computer vision or machine learning algorithms. It involves a series of operations that aim to correct, filter, normalize, or enhance the images to improve their suitability for subsequent analysis. Digital image pre-processing involves several crucial steps to enhance image quality and accuracy for further analysis. These include geometric correction, radiometric correction, and orthorectification. Geometric correction addresses distortions caused by the Earth's curvature, sensor angle, and terrain, ensuring accurate spatial alignment. Radiometric correction compensates for variations in lighting, atmospheric conditions, and sensor inconsistencies, ensuring accurate spectral representation. Orthorectification, a type of geometric correction, uses digital elevation models (DEMs) to remove parallax errors caused by relief, producing a geometrically accurate and map-aligned image.
Radiometric Correction
Radiometric correction of remotely sensed data typically involves the processing of digital images to improve the fidelity of the brightness value magnitudes (as opposed to geometric correction, which involves improving the fidelity of relative spatial or absolute locational aspects of image brightness values). Each element on the data arrays contains a digital number (DN) proportionate to the electric charge received at capture time. However, the values do not necessarily express the objects’ actual radiation or reflectance due to several error interferences, either by the sensor operation or environmental factors referred to as noise. The main purpose of applying radiometric corrections is to reduce the influence of errors or inconsistencies in image brightness values that may limit one’s ability to interpret or quantitatively process and analyze digital remotely sensed images. Radiometric correction is particularly important when images will be compared across time to investigate changes or when they will be used for spectral analyses, such as computing plant health indices.
Geometric Correction
Geometric correction is the process of compensating for both systematic and random distortions in the size and locations of image pixels. The goal of geometric correction is to ultimately produce a corrected image with a high level of geometric integrity in which the data are in their proper x, y, and sometimes z locations. Any remote sensing image, regardless of whether it is acquired by a multispectral scanner on board a satellite, a photographic system in an aircraft, or any other platform/sensor combination, will have various geometric distortions.
Selection of Geometric Correction Method
In digital image pre-processing, selecting the appropriate geometric correction method involves identifying the type and magnitude of geometric distortions in the image and choosing a method that accurately corrects them. Common methods include image-to-map rectification, image-to-image registration, and the use of ground control points (GCPs). The choice depends on the desired outcome, such as georeferencing, overlaying multiple images, or integrating the image with other spatial data. Choosing the right geometric correction method and GCPs significantly impacts the accuracy of the final corrected image.
Spatial Interpolation
A key part of this georeferencing process is spatial interpolation, the process of estimating the pixel values at newly transformed coordinates after applying a geometric transformation. Because new pixel positions often do not align exactly with the original grid, interpolation determines what the pixel values should be.
Intensity Interpolation (Resampling)
Once the mapping transformation has been determined, a final procedure called resampling is employed. Resampling matches the coordinates of image pixels to their real-world coordinates and writes a new image on a pixel-by-pixel basis. Since the grid of pixels in the source image rarely matches the grid for the reference image, the pixels are resampled so that new data file values for the output file can be calculated (Figure 15.6). This process involves the extraction of a brightness value from a location in the input image and its reallocation in the appropriate coordinate location in the rectified output image. Different resampling methods can be used in the rectification methods. The three most common resampling methods are nearest neighbor assignment, bilinear interpolation, and cubic convolution.
Georeferencing
Georeferencing is the process of taking digital images or maps and assigning them real-world coordinates. By doing so, these images can be accurately overlaid onto existing spatial data, allowing for a seamless integration of diverse datasets. Historically, imagery was often misaligned due to the limitations in capturing and storing spatial data.
Orthorectification
mage data acquired by airborne systems is affected by systematic sensor and platform-induced geometry errors, which introduce terrain distortions when the sensor is not pointing directly at the nadir location of the sensor. Orthorectification aims to create a final product whereby every pixel in the image is depicted as if it were viewed at a nadir (or directly overhead), thereby removing the effects of hills, valleys, etc., on the data (Figure 15.7).
Image Registration
Image registration in digital image analysis is the process of geometrically aligning two or more images of the same scene, or different images of the same scene taken at different times, from different viewpoints, or with different sensors.
Image Mosaicing
Georeferenced (or orthorectified) images with different spatial resolutions, brightness histograms, and sizes can be digitally mosaicked (i.e., join two or more images with overlapping regions) with advanced image processing software (Figure 15.8). Image mosaicing is still one of the hot subjects, especially for remote sensing applications.
Click on the following topics for more information on digital image processing of remote-sensed data.
