Visualizing, processing, analyzing, and managing imagery and raster data are an important part of GIS. Information contained in typical land base maps and GIS layers is derived from imagery and raster data. Existing maps and layers are revised based on updated imagery.
ArcGIS includes capabilities for visualizing, managing, processing, and analyzing imagery and raster data. These imagery and remote sensing capabilities are extended in the more advanced image analysis capabilities in the ArcGIS Image Analyst extension.
Imagery and remote sensing defined
Images are raster data that represent measurements of reflected or emitted electromagnetic energy captured by a sensor on a drone, airplane, or satellite. Other types of imagery and raster data include scientific measurements of a location's properties, such as temperature and salinity at different water depths, elevation models, and seismic surveys.
Remote sensing extracts meaningful information from imagery. It applies image processing techniques to identify and extract various types of information about features of interest, such as vegetation type and health, type of urban development, and identifying and analyzing trends in detected objects and phenomena. Proximity and neighborhood analysis, seasonality, physical and cultural geography, vegetation growth phenomenology, climate, and other physical and environmental factors are considered when defining the type and timing of image acquisition. Remote sensing is the embodiment of The Science of Where™.
The tools, functions, and capabilities in ArcGIS extract spatial and spectral information from imagery using scientific and image processing methodologies. While the processes are standardized and optimized using advanced image processing techniques, workflows and results are often unique based on scientific remote sensing techniques. For example, the relative health of specific vegetation species can be monitored and analyzed over a growth cycle.
Imagery and remote sensing provide functions, tools, and capabilities to perform the following:
- Correct, calibrate, and standardize imagery for integration into workflows
- Photogrammetric correction of remotely sensed imagery from drones, aircraft, and satellites
- Image interpretation, processing, and exploitation
- Multispectral and multidimensional raster analysis and processing
- Create information products from imagery
- Advanced raster and image analysis workflows for machine learning and feature extraction
Manage imagery and raster datasets
AllSource provides a collection of tools to manage and serve large collections of imagery. The core components of image management are mosaic datasets and raster products. Traditionally, a mosaic is created when you combine a collection of images to create one seamless image. The mosaic dataset does that, but it also uses image attributes, such as acquisition date, cloud cover, or spatial resolution, to help you manage a collection of imagery, even one composed of overlapping images or images with varying spatial resolutions. Instead of creating files when you want to change which images are displayed, or enhance the images, the mosaic dataset processes the imagery in-memory, allowing it to quickly display an imagery of interest, regardless of the size of the collection.
Most imagery includes a metadata file that describes the sensor and the conditions under which the image was acquired. AllSource identifies the structure of imagery from most commonly used sensors, and reads the metadata file to compile all the bands into layers that can display various combinations of multispectral bands, pan sharpened imagery, or thermal bands, depending on the capabilities of the sensor. Similar to a mosaic dataset, band composition occurs in-memory, so it displays quickly without creating files.
Process and analyze imagery
AllSource contains tools and raster functions that work with imagery and raster data. Regardless of whether pixel-based data is an image from a satellite, a drone or aerial sensor, a raster dataset, or a DEM, there are many ways that you can work with this data when performing analysis.
Traditionally, image and raster analysis is performed with standard and advanced geoprocessing tools. The geoprocessing tools, Python scripting, and ModelBuilder work together to provide processing and automation of data management operations in the geoprocessing framework. While these geoprocessing tools are used in many workflows, AllSource provides another approach to image and raster analysis that processes and displays results immediately as you pan and zoom imagery.
Distributed raster analysis uses Portal for ArcGIS to improve the performance of raster processing by distributing the workload to ArcGIS Image Server.
Raster functions
Using raster functions is one way to process and analyze mosaic datasets and rasters in ArcGIS. This approach processes imagery and raster datasets on the fly as they are accessed, and displays results immediately in a map display. Raster functions work in-memory and avoid the time and overhead of reading and writing intermediate files to disk. A suite of raster functions is available out of the box. You can use individual raster functions, or chain multiple functions together in a custom raster function template (RFT) using the Function Editor. The Function Editor is a visual programming interface for building, testing, and saving image and raster analysis workflows. You can save both results and RFTs as a file.
Distributed processing
AllSource allows you to use Portal for ArcGIS to perform distributed server-side processing on imagery and raster data. This technology allows you to improve the performance of raster processing by processing data in a distributed manner using ArcGIS Image Server configured for Raster Analysis. You can use geoprocessing tools and models, and raster function processing capabilities, to perform processing in this way using an on-premises portal. It accepts local files and items from your portal as input and creates output in the same portal.
You can also access, process, and publish hosted imagery and rasters in ArcGIS Image for ArcGIS Online using Portal for ArcGIS. All the image processing capabilities in AllSource are available for processing hosted imagery layers from ArcGIS Image for ArcGIS Online, and publishing results to your portal account.
Visualize imagery
From the contextual tab for an image layer type, you can access the functionality to preprocess imagery, alter the display or the way the data is rendered, enhance imagery, change the symbology, and compare one layer with another. The process for working with mosaic datasets is generally the same as it is for a single image, except that you can specify that operations be performed on specific items comprising the mosaic dataset. Additionally, image items displayed in a mosaic dataset are defined by metadata properties.
AllSource allows you to display imagery in 2D and 3D space. Multiple 2D and 3D views can be linked so you can work with different data types and different views of the same area as you roam and zoom the data. This capability results in processing efficiencies and better analytical and visual analysis.
Oblique imagery can be displayed in its natural orientation in a native image coordinate system (rows and columns) rather than a map orientation with north at the top. This is helpful for image interpretation.
You can view and work with imagery in stereoscopic mode. Choose from a variety of active or passive stereoscopic setups. See Stereo mapping in AllSource for more information.
Ortho Mapping
Ortho Mapping in AllSource provides tools, capabilities, and guided workflows to perform rigorous orthorectification of drone, aerial, and satellite imagery. These include project setup, definition of control points, tie points and check points, block adjustment, and elevation and orthomosaic product generation. Reports and visual diagrams provide analytical and quantitative information for refinement to achieve optimum orthorectification results.
To learn more about orthorectification, see Introduction to ortho mapping.
Stereo mapping
AllSource stereo mapping in Image Analyst allows you to view and work with stereo pairs and collect 3D features. You can collect 3D point, line, and polygon features for various workflows such as parcel editing, creating simple 3D buildings, and measuring distances and heights.
Deep learning
Deep learning tools detect features in imagery using multiple layers in neural networks in which each layer is capable of extracting one or more unique features from the image. These tools use GPU processing to perform the analysis in a timely manner.
For more information about deep learning using imagery and raster data, see Deep learning in Raster Analysis.
Classification
Image classification allows you to extract classes from multispectral imagery. The raster resulting from image classification can be used to create thematic maps. You can classify image pixels or segmented objects using parametric or machine learning classifiers. You can use the Image Classification Wizard to guide you through the classification process. To perform only a portion of the classification workflow, you can use the various image classification tools.
Multidimensional analysis
Multidimensional analysis tools and capabilities allow you to perform and visualize complex analysis on multidimensional raster data to explore scientific trends and anomalies. Multidimensional data represents geospatial data captured at multiple times and multiple depths or heights. These data types are commonly used in atmospheric, oceanographic, and earth sciences. You can capture multidimensional raster data by satellite observations in which data is collected at certain time intervals or generated from numerical models in which data is aggregated, interpolated, or simulated from other data sources.
For more information about multidimensional data, see Multidimensional raster data.
Georeferencing
Georeferencing imagery defines its correct geographic location so it can be viewed, queried, and analyzed with other geographic data. To georeference data, create control points that tie the raster to real geographic coordinates. The tools to georeference raster data are on the Georeference tab.
Imagery and raster charts
Imagery charts are designed to interactively process and display multispectral and multidimensional raster data for image analysis and assessment. For example, you can delineate an area of interest on an image, and plot the spectral profile of specific ground features such as human-made development, vegetation types and health, and other features to support specific applications with quantitative data and information. Imagery and raster chart types include spectral profile, temporal profile, image scatter plot, dimension profile, image histogram, and bar. You can use the Pixel Time Series Change Explorer to identify changes in a single pixel value over time using any change detection method.
