The 2012 Connecticut aerial photography project produced 1' (.30m) GSD multispectral digital orthoimagery for the entire state. An orthoimage is remotely sensed image data in which displacement of features in the image caused by terrain relief and sensor orientation have been mathematically removed, thus it combines the image characteristics of a photograph with the geometric qualities of a map.
The digital orthoimagery is both high resolution and spatially accurate, and can serve as a raster base image for mapping projects as well as a data source to support the derivation of new raster and vector data sets.
Digital aerial imagery was obtained in the spring of 2012 using a large format Z/I Digital Mapping Camera system (DMC) equipped with Airborne GPS/IMU. A total of 59 flight lines with 6,287 frames were acquired under a single task order, in multispectral (RGB and NIR) 12 bits per band format. The imagery was acquired with a 4.7244" (120 m/m) focal length at an altitude of 10,000' above mean terrain, to yield a raw pixel resolution of 1' (.30m) suitable for photogrammetric mapping and orthophoto production. The leaf-off imagery was collected in the late spring of 2012 under conditions free from clouds and cloud shadows, smoke, fog, haze, light streaks, snow, ice on water bodies, foliage, flooding, and excessive soil moisture. The sun angle threshold was 30 degrees. The imagery consisted of panchromatic, blue, green, red and near infrared bands. The three color bands and near infrared bands were pan sharpened and archived as frame imagery. All 4 bands were used in the orthophoto production.
The United States Geological Survey shall not be held liable for any errors in this data. This includes errors of omission, errors of commission, content errors, and relative and positional accuracy errors in the data. This data should not be construed to be a legal document. Primary sources from which this data was compiled must be consulted for verification of information contained in this data. This data is in the public domain, and may not be resold.
1111 Country Club Road
Email is preferred.
Orthophotos were produced for USGS by Photo Science under the USGS Geospatial Production and Services Contract, Contract Number G10PC00026, Task Order G12PD00166. They were paid for by the State of Connecticut Department of Emergency Services and Public Protection and the State of Connecticut Department of Transportation.
Radiometry was verified by visual inspection of the digital orthophotos. Slight systematic radiometric differences may exist between adjacent orthoimage files; these are due primarily to differences in source image capture dates and sun angles along flight lines. These differences can be observed in an image's general lightness or darkness when it is compared to adjacent orthoimage file coverages. Tonal balancing was performed over a group of images during the mosaicking process in order to lighten or darken adjacent images for better color tone matching.
All GeoTIFF tagged data and image file sizes were validated using commercial GIS software to ensure proper loading before being archived. This validation procedure ensured correct physical format and field values for tagged elements. Seam lines and tile edges were visually inspected.
Photography was flown in the spring of 2012, under conditions free from clouds and cloud shadows, smoke, haze, light streaks, snow, foliage, flooding, and excessive soil moisture, and when the sun angle was no less than 30 degrees. To ensure complete coverage, photography was collected for all tiles that intersected a 0.25 mile buffer around the entire state. Orthoimages were visually inspected for completeness to ensure that no gaps or image misplacements exist within and between adjacent images.
This 0.30m Ground Sample Distance (GSD) digital ortho imagery was compiled to meet a 1.52 meter (5') horizontal accuracy at 95% confidence level(0.88-meters Root Mean Squared (RMSE) Error XY (0.62 meter RMSE X or Y) based on NSSDA testing guidelines. This accuracy requirement is not applicable in areas where the ground is obscured on the aerial imagery by foliage, prevalent smoke, or dense shadow.
There is no vertical component for orthophotos.
The digital aerial imagery was obtained using a large format Zeiss/Intergraph Digital Mapping Camera (DMC). Aerial imagery was supplemented with the simultaneous acquisition of airborne GPS/IMU data, which captured the ground coordinate for the nadir point of each photograph. Aerial imagery was exposed at an altitude of 10,000' above mean terrain yielding an approximate GSD of 1 foot. A minimum of 60% forward gain and 30% sidelap was collected to ensure full stereo coverage. The aerial images are used in the production of orthophotos.
Targeted ground control was collected for use in the Aerotriangulation process. One hundred (100) photo identifiable ground control points were selected and surveyed to adequately cover the entire state. Details of each point can be found in the full control report.
The first primary source used to support orthorectification was USGS's LiDAR for the Northeast project for two smaller coastal areas located in the southeastern and southwestern corners of the State. LiDAR data for these two areas was collected in the spring and fall of 2011. Photo Science was the assigned USGS Contractor for this Project and is familiar with the associated technical specification (2 meter NPS) and accuracies (15cm RMSEz in open terrain) Photo Science will utilize the bare earth LiDAR point class that it has in its archive to support the ortho rectification process. The second primary LiDAR derived DEM source covers the eastern half and northwestern corner of the State. The NRCS Phase 1 LiDAR dataset, commissioned by the NRCS and contracted through the USACE, St. Louis District, was reportedly acquired in the fall of 2010 during leaf off conditions at a 0.7 meter ground sample distance achieving a vertical accuracy of 9.2cm RMSEz to support the development of one (1) foot topographic contours. The NRCS Phase 2 LiDAR dataset was reportedly acquired to the same specifications as Phase 1 in the fall of 2011 Photo Science will utilize the bare earth LiDAR point class from these two NRCS projects to support the ortho rectification process. The third primary LIDAR derived DEM data source used covers the Quinnipiac Watershed (~443 sq. miles) and was acquired by Photo Science in 2010 to FEMA Region 1 via the STARR Production Technical Services Contract in support of the Risk Map Program. Photo Science developed classified, bare earth, .LAS LiDAR data at a 1-meter nominal post spacing with a 24.5 cm vertical accuracy at 95% confidence interval. Photo Science will utilize the bare earth LiDAR point class that it has in its archive to support the ortho rectification process. The fourth primary LiDAR derived DEM source used by Photo Science for any areas not covered by more recent and/or more accurate LiDAR (approx 25% of State) is the statewide dataset consists of LiDAR-derived data products generated from Connecticut's 2000 statewide LiDAR coverage. Derived products include tiled point clouds, TINs, 10 foot gridded DEM, and contour shape files. The dataset that Photo Science used a thinned, ground point dataset (txt format) at nominal 20' postings in USGS quarter-quarter quadrangle format to support orthorectification. All derived products are projected in Connecticut State Plane, NAD83, Feet. Photo Science has been unable to verify the published or tested accuracy and fidelity of the 2000 LiDAR data.. Because the 2000 CT LiDAR data is 12 years old, it was anticipated that changes in terrain will introduce a higher degree of localized imagery warping/smearing as compared to the other newer LiDAR sources described above. Photo Science made reasonable effort to identify these errors and correct excessive image smearing/warping during the ortho finishing phase. Secondary LiDAR data sources that Photo Science used include Connecticut River and Coastal LiDAR datasets. It is understood that these sources were collected by USACE and/or NOAA. Little is known about their data characteristics, accuracy and age. Photo Science did not update any of the LiDAR DEM sources Additionally, the use of LiDAR datasets not possessing adequate vertical accuracies to support this Task Order's orthophoto horizontal accuracies may impact the final orthophoto accuracies in the affected areas. Photo Science will immediately notify USGS if it determines that orthophoto qualitative or quantitative requirements are being negatively impacted due to LiDAR source issues.
Exterior Orientation (EO) Parameters (x,y,z,omega,phi,kappa) were produced to provide precise photocenter locations for each frame of imagery. Aerotriangulation was necessary to achieve the accuracy requirements of the project.
SoftCopy Aerotriangulation is a computer process of densifying the ground control network from locations in a few of the digital images to numerous points in each digital image that is to be used in the mapping process. This process is dependent on the location and coordinate positions of the ABGPS and GPS photo control points as well as a precise camera calibration for the DMC camera used to capture the images.
The tile definition was used to cut orthorectified imagery into manageable, usable images in State Plane CT projection. The UTM deliverables defined by this tile definition were modeled after the U.S. National Grid and references NAD83 UTM projection, units of Meters. Each tile is 1500 X 1500 meters.
This preliminary delivery of images were the product of the processed stereo imagery frames and the Aerotriangulation solution. Digital imagery was produced in TIF file format, 4-band, 8-bit and delivered to the State as stereo pairs. Unrectified stereo frame imagery of the missions was delivered in a format suitable for viewing with ERDAS Stereo Analyst Extension of ArcGIS.
Orthorectified images were geometrically corrected to achieve a uniform scale. Each frame was adjusted for topographic relief, lens distortion and camera tilt.
The tiled orthophoto was the result of mosaicking orthorectified imagery and using the tile definition to cut the data into manageable, usable orthophotos. 6103 complete ortho tiles for delivery to USGS and 6131 State Plane 5000' x 5000' ortho tiles (some of which are partial) were created.
One hundred (100) photo identifiable control locations were selected and surveyed to adequately cover the entire state. Details of each point can be found in the full control report. Accuracies of 10cm or better (horizontally and vertically) were required.
New digital aerial imagery was obtained in the spring of 2012 using a large format Z/I Digital Mapping Camera system (DMC) equipped with Airborne GPS/IMU. A total of 59 flight lines with 6,287 frames were collected in multispectral (RGB and NIR) 12 bits per band format. The imagery was acquired with a 4.7244" (120 m/m) focal length at an altitude of 10,000' above mean terrain, to yield a raw pixel resolution of 1' (.30 m) suitable for photogrammetric mapping and orthophoto production. The imagery was collected during leaf off conditions in the spring of 2012 under conditions free from clouds and cloud shadows, smoke, fog, haze, light streaks, snow, ice on water bodies, foliage, flooding, and excessive soil moisture. The sun angle threshold was 30 degrees. The imagery consisted of panchromatic, blue, green, red and near infrared bands. The three color bands and near infrared bands were pan sharpened and archived as frame imagery.
Imagery and AGPS/IMU data is downloaded from the hard drive on the plane to the hard drive on the ground. AGPS/IMU mission data is processed together with continuously collected ground-based CORS GPS base station data in forward and reverse directions. This precisely determines the aerial camera's position and orientation in the terrain (project) coordinate system and allows for correct orientation of the imagery.
The EOFILE, raw frames, and control points were input into ISAT in order to complete the SoftCopy Aerotriangulation, described in ATSOL source contribution, and assembled into a block. A second degree block adjustment is run along with the full least squares bundle adjustment. Once the last set of blunders and adjustments are complete, a final block bundle adjustment is run to produce the final solution. The final adjustment of the block is accomplished by using a rigorous simultaneous least squares bundle adjustment. The general procedure is to remove all blunders from the data using automatic blunder detection. The bundle adjustment is then run with minimal ground control to test the photogrammetric measurements for consistency. Next, the full ground control data set, including ABGPS data, is added to the adjustment. The horizontal control is then tightened and the effect on the vertical control and the photogrammetric residuals are inspected. The final adjustment makes sure that all of the measurements are in balance with each other and properly represent the actual conditions.
The rectification process required as input a digital elevation model (LiDAR_DEM), imagery (IMG1) and soft copy aerotriangulation solution (ATSOL). The orthorectified imagery was produced using Intergraph Orthopro software. It used ImageStation Photogrammetric Manager (ISPM) for photogrammetry project setup, file management, and importing triangulation data from a user defined ASCII file (ATSOL). The ortho project records all the operations, including all the parameters that are defined through the OrthoPro workflow, such as the project area coordinates, pixel size, rotation angle, ISPM project and its unorthorectified source images, product areas (TLDEF) orthorectification settings, seamlines and so on. The results are single frame images that have yet to be mosaicked into manageable tiled orthos. The single frame orthos were checked for accuracy against the surveyed ground control before further image editing.
Orthorectified imagery was mosaicked, locally color-balanced and cut to the tile definition boundaries (TLDEF) using Intergraph Orthopro software. Tiled orthophotos then went through a rigorous manual QC process to evaluate for remaining hotspots (sun reflectance over water), tone quality, color balance and the feathering area along seamlines. Any imperfections at this point were manually edited. The tiled images were then converted to the required file format (GeoTIFF) and transferred to external hard drives for delivery to USGS.
Data contributed to CTECO by the Connecticut Department of Emergency Services and Public Protection. Metadata reviewed and nominally updated with a local point of contact, access and use constraints, and distribution information.
1066 Saybrook Rd.
PO Box 70
These digital orthophotos are composed of 4 bands or channels of information. Each band comprises a grid of pixels containing digital numbers ranging from 0-255, and representing colors in the red, green, blue (RGB) or near infrared (NIR) portions of the electromagnetic spectrum.
1066 Saybrook Rd.
PO Box 70
Email is preferred.
Although these data have been used by the State of Connecticut, no warranty, expressed or implied, is made by the State of Connecticut as to the accuracy of these data and or related materials. The act of distribution shall not constitute any such warranty, and no responsibility is assumed by the State of Connecticut in the use of these data or related materials. The user assumes the entire risk related to the use of these data. Once the data is distributed to the user, modifications made to the data by the user should be noted in the metadata. Digital data provided by USGS represent the efforts of the contributing agencies to record information from the cited source materials. USGS and the cooperating agencies make no claim as to the validity or reliability or to any implied uses of these data. These data are in the public domain, and may not be resold.
Tiles as delivered by USGS
Connecticut Town Mosaics
The data distributor does not provide custom GIS analysis or mapping services. Data is available in a standard format and may be converted to other formats, projections, coordinate systems, or selected for specific geographic regions by the party receiving the data.
1400 Independence Road
PO Box 70
Email is preferred.