The IGN CAMv2 System

Abstract

After 10 years of pioneering aerial digital camera development, IGN’s CAMv2 project began in 2006 with the main goal of upgrading the then current system. This had been used for research applications since 1996 in more than six different configurations and in production since 2003 configured with four coloured channels. The new system is still highly modular, enabling combinations of several “basic components”, each consisting of one digital camera head with its storage and control unit. The new camera head was developed around the Kodak KAF‐39 000, 39‐megapixel array sensor (7216 × 5412 pixels). The system is also very versatile thanks to the possible choice of more than 10 different lenses (from 35 mm up to 180 mm focal length) and interchangeable spectral filters. The mechanical and electronic designs have been reduced in size so as to permit many different configurations including multiple camera heads on the same gyro‐stabilised mount. The combination of the performance of camera heads and control and storage units allows a frame rate of one image per 2 s with storage redundancy and 1 s without, leading to a minimum ground sample distance (GSD) of about 9 cm with 60% overlap at an airspeed of 100 m/s. In order to use the new system to best advantage, the adjustment and calibration processes had to be improved. During the summer of 2009, three systems configured for colour imagery, each using four nadir‐viewing camera heads (RGB and NIR), were used to acquire images for IGN production work. In January 2009, an eight‐camera‐head configuration was tested in order to achieve 155‐megapixel pan‐sharpened images with a pan‐sharpening ratio of 2 × 2, and a final swath width of about 14 400 pixels. At that time this configuration needed two windows in the aircraft, but it is now installed on a single, bigger mount.

Feature matching for multi-epoch historical aerial images

“…We have at our disposal three sets of images acquired in 1971, 1981 and 2015. The epoch 2015 was acquired with the IGN's digital metric camera [63], and is treated as the ground truth (GT) during our processing (in other words, the 2015 image orientations are fixed during BBA). The area exhibits changes in scene appearance in the 44-year period.…”

Detection of individual trees in urban alignment from airborne data and contextual information: A marked point process approach

“…About the spatial resolution, the VNIR camera data are acquired with a spatial resolution of 0.4 m and 0.8 m across and along tracks, respectively. In order to build a DSM, the French Mapping Agency CAMv2 system was used (Souchon et al, 2010) for performing stereoscopic acquisitions with an overlap of 80%. A vector layer of roads derived from the OSM database is used and identified as "GIS data" in the next sections.…”

Hyperspectral and Radar Airborne Imagery over Controlled Release of Oil at Sea

“…The optronic sensors, installed in the SETHI pods, are composed of one panchromatic camera (CamV2 [ 13 ], developed by the French National Mapping Agency, IGN (Institut Geographique National)) and two HySpex VNIR1600 and SWIR320m-e hyperspectral cameras built by NEO (Norsk Elektro Optikk) [ 14 ]. The main parameters of optical sensors operated during the NOFO oil-on-water exercise are given Table 2 .…”