MAC URISA 2016

The combination of vertical with oblique image cameras in only one mechanical set is not a new invention. As early as 1900 Scheimflug invented a camera with 8 oblique lenses and one nadirlens. Since then, different setups have appeared in the industry with some success in their application. Lately, the advent of new POS technology as well as the new digital CCD chips being used in the commerciallyavailable mid-size digital cameras has revived this sort of image data acquisition that allows the 3D-reconstruction of the object space from aerial images. Satellite imagery is also used with the same purpose. Applications are countless, nearly all being related to 3D-modeling of engineering objects and the 3D mapping is that objects can be classified and joined in layers that, along with the incorporation of their attributes, will form the basis of a multitude of 3D GIS applications. The present paper is a summary review of the latest acquisition systems, their advantages and disadvantages, complex geo-spatial tasks to be followed for a better review of cases studies and projects that are presented and analyzed from both visual and statistical viewpoints.

The U.S. Geological Survey, with support from the New Jersey Department of Environmental Protection (NJDEP), has undertaken a study to develop an approach to improve the locational accuracy and precision of the New Jersey stream network using lidar-derived Digital Elevation Models (DEMs). Regional terrestrial lidar collections in northern New Jersey during 2006-08 at quality level 3 and 2014 at quality level 2 yielded bare earth DEMs with cell sizes of 3 meters and 1 meter, respectively. The DEMs were hydro-enforced to remove surface features at stream crossings such as road culverts and bridges, permitting unobstructed flow of water across the landscape. A bottom hat morphologic filter was applied to the enforced DEM datasets to derive local-scale hydrography in three 12-digit Watershed Boundary Dataset basins in different physiographic provinces in the Raritan River Basin. The approach applies raster processing filters to a DEM following a series of steps described in Cho and others (2011) and Rodriguez (2002). A low pass filter is created from the DEM, followed by a focal maximum, and a minimum of the maximum or closing. The closing is subtracted from the low pass filter creating a bottom hat filter. The process steps include filling sinks and computing flow direction, flow accumulation, and a weighted flow accumulation. The resulting raster is converted to vector, delineating the stream lines. Stream lines derived in watersheds in the Highlands, Piedmont, and Coastal Plain Physiographic Provinces in the Raritan River basin were compared to 1:2,400 scale National Hydrography Dataset (NHD) stream lines. The local-resolution NHD in New Jersey was digitized by the NJDEP from visual inspection of 1:2,400 scale color orthoimagery acquired in 2002. These datasets provided a unique opportunity to compare stream lines derived from lidar data to local-resolution hydrography. Visual inspection of, and spatial comparison statistics for, the derived stream lines and the NHD were used to assess the utility of the processing methods in each physiographic province. Comparison of the stream network derived from the 3-meter DEM indicates that 74 percent of the NHD stream lines are within 10 meters and 67 percent are within 5 meters of the derived streams in the Highlands Physiographic Province. In the Piedmont Physiographic Province, 57 percent of the NHD stream lines are within 10 meters, and 44 percent are within 5 meters. Forty-two percent of the NHD stream lines fall within 10 meters of the derived stream network in the Coastal Plain Physiographic Province, and 30 percent are within 5 meters. Field inspection at 75 verification sites provided ancillary information, such as channel type, width, flow characteristics, and type of control structures, further documenting the accuracy of the approach. 

Pictometry is a leading provider of geo-referenced aerial oblique and orthogonal image libraries and related software. Pictometry technologies are widely used by GIS, assessing, emergency services, and planning professionals around the country.