Presentations in this session will show how advanced photogrammetric techniques are applied to imagery acquired with a variety of optical sensors to produce high resolution/high-accuracy mapping and information products.
11:00 – 11:15 AM – Is Stereo Dead?
Stereo photogrammetry was the foundation of the mapping industry for close to a century. Nearly all maps produced from 1930 to 2000 were created by stereo compilation. There was heavy investment in instrumentation as the technology developed and transition from mechanical, to analytical, to fully digital. Today many consider stereo as being reduced to a mathematical formula implemented by multiple software vendors with variations on viewing hardware for how to drive separate images to the right and left eye. Many of the traditional stereo mapping workflows have been replaced by newer technologies such as Lidar and Reality Mapping or automated to the point that viewing stereo is not considered necessary. Proportional to the size of the mapping industry the use of stereo has reduced to a fraction of the past. Should this be the case? The presentation provides an explanation of the advantages and disadvantages of stereo viewing for interpretation and compilation of features and terrain breaklines, so that mapping practitioners can better determine where stereo viewing is advantageous.
Peter Becker, Esri
11:15 – 11:30 AM – Unlocking the Power of Non-Traditional Imagery in GIS
Photo collection from drones, smartphones, and 360° cameras is revolutionizing asset inspection and management. However, integrating these non-traditional image types into GIS systems can be challenging. Two critical factors are at play. First, the widespread availability of drones and cameras today generates an overwhelming volume of visual data. Second, while many tools exist for working with orthographic imagery, few allow perspective images to be integrated into a GIS environment.  Perspective imagery within GIS exists between two extremes. On one end, geolocated photos provide basic location data (lat/long) but lack precision in direction, angle, and field of view, making them informative but limited. On the other end, photogrammetry requires extensive preplanning, multiple inputs, and sophisticated software to create highly precise 3D models. With Oriented Imagery, which can be thought of as “Photogrammetry Lite,”in which a single picture contains sufficient metadata to be made actionable in a mapping environment. Fully actionable in this case means not just knowing where a photo was taken, but enough additional data (pan, tilt, roll, FOV, etc) to identify where every pixel in that photo exists in the real world.  By combining readily available sensors and leveraging innovative Oriented Imagery technologies, drone pilots and field crews can now create, edit, and share actionable content directly within GIS environments. This approach significantly reduces the time and effort needed to convert raw visual data into actionable GIS content. Allowing field teams to directly handle and georeference imagery, organizations can streamline workflows, improve data accuracy, and reduce dependency on specialized GIS analysts. Empowering field personnel not only enhances operational efficiency but also ensures that the most knowledgeable individuals about the data are involved in its processing and curation. For teams in the utility and government sectors, the ability to process more imagery with ease and reference it with existing GIS data offers substantial advantages. These organizations can fully leverage the potential of non-traditional imagery, resulting in more efficient asset management, improved data quality, and quicker decision-making processes.
Nick Markwardt, AerialSphere
11:30 – 11:45 AM – Comparing Nonlinear Least Squares and Closed Form Solutions of the 3D Conformal Coordinate Transformation
The 3D conformal coordinate transformation has applications in photogrammetry, lidar, robotics, and geodesy. It is also known as absolute orientation and the seven-parameter transformation. For this transformation, one must express the coordinates of a set of 3D points known in one coordinate system relative to another coordiante system. One means of finding the relationship between two coordinate systems is to measure the coordinates of a set of points common to both the coordinate systems. Once the relationship between the two coordinate systems—the transformation—is solved, the whole of the point data in one system can be transformed to the other. The transformation from an initial to a target system is commonly expressed as seven parameters: A uniform scale factor, three sequential rotation angles, and three translations. This transformation preserves the shape of the data being transformed, i.e., no skew, non-uniform scaling, perspective warp, etc. is introduced. There are a number of means of solving for the transformation, each with their advantages and shortcomings. The purpose of this talk is to explore these pros and cons of (1) the closed-form solution using unit quaternions presented by BK Horn and (2) the nonlinear least squares approach commonly taught in photogrammety courses. The goal is to provide guidance to researchers and practitioners who wish to examine the efficacy of any software packages which perform the 3D conformal transformation, or to develop their own implementations of the transformation.
Andrew Lassiter, University of Florida
11:45 – 12:00 PM – Professional Grade Crewed Multi-Head Aerial Camera Systems in Direct Georeferencing Environment
Photogrammetric Cameras for Aerial surveys have evolved since WWII… Today, a typical aerial survey camera system consists of multiple cameras acquiring nadir and oblique images as well as RGB and NIR images at the same moment of exposure from those multiple cameras embedded in the system. This enables a variety of high precision mapping applications including city, state, country, and continent mapping projects. Furthermore, these systems typically contain several embedded sophisticated sensors including GNSS/IMU sensors that allow for producing a variety of mapping products using Direct Georeferencing (DG) technology without the need for the traditional photogrammetric workflow that heavily relies on Ground Control Points (GCP’s) … The DG approach has many advantages to it since reduces technical challenges and logistical constraints and improves the overall operational procedures to deliver highly accurate mapping products. However, the DG approach relies on multiple streams of data from different nadir/oblique cameras as well GNSS/inertial data. Optimizing data acquisition and processing is a key factor for the success of the direct georeferencing method. This technical pedagogical presentation introduces the main elements of managing large aerial survey projects using Multi-Head Camera Systems in a DG environment where flight planning, data acquisition, data processing, quality assurance and quality control are addressed in detail. Multiple datasets acquired in the US and Canada are used for this presentation.  This presentation is intended to educate the on the optimal use of professional grade multi-head camera systems for aerial surveys. This is not a sales or marketing presentation; it is a technical presentation focused on education and promoting technology but not products … Accuracy assessment will be addressed using different data acquisition and processing approaches along with their associated statistical measures.
Mohamed Mostafa, Trimble Applanix
12:00 – 12:15 PM – Terrestrial Scanning for Construction, Engineering, and Surveying
Photogrammetry extends beyond just aerial platforms. Modern mobile devices are equipped with high-quality cameras and often include LiDAR and depth sensors. When paired with precise GNSS systems and visual positioning methods, these devices can efficiently gather accurate data for utilities, infrastructure, inspections, as-built surveys, and more. This presentation will offer an overview of terrestrial photogrammetry technology, highlighting best practices, accuracy, and applications in construction, engineering, and surveying. We will demonstrate how leveraging these accessible devices enables fast and accurate digital infrastructure modeling, allowing us to keep digital twins up to date without relying on costly and time-consuming alternatives.
Tom Cerchiara, Pix4D
12:15 – 12:30 PM – The Smear Phenomenon in High-resolution Orthorectified Satellite Imagery
You may have noticed an artifact in processed high-resolution imagery in which pixels appear to be smeared across the surface, similar to paint across a canvas. It is an unfortunate side effect from the orthorectification processing of some high-resolution images, whether collected from satellite or aerial platforms, that needs to be avoided or dealt with if possible. This presentation explores the cause of this effect, why and when it happens, and some methods for mitigating it. If the satellite look angle and azimuth are known, smear can easily be identified and masked automatically. It is a function of the satellite viewshed and vertical features on the Earth’s surface causing obstructions, and how the imagery is stretched over the digital surface used in orthorectification. We can predict when smear will happen in imagery prior to production, and that knowledge may inform certain processing decisions such as what resolution of DEM to use, or which image in an archive to select. There may be some circumstances when smear is unavoidable in a product, but being ahead of it and informing the customer by explaining and mapping it in the image product is better than allowing a customer to discover it on their own.
Francois Smith, Maxar Intelligence
