Florin Savopol 
The most difficult problem was the synchronization. A synchronization signal generator was build following the 
inge of the technical specifications from the Camera manufacturer. Initially, a significant “jitter” of up to 5 to 10 pixels was 
present when using the analog mode and, in the digital mode, the image from the second camera was unusable. The 
experiment was repeated using different configurations for the synchronization signals. In digital mode it was still 
impossible to capture useful images from the second camera. The results were improved in analog mode using two HP 
signal generators for the vertical and horizontal synchronization. The images from the first cameras were distortion free 
and those from the second one had generally a jitter of about 2 to 4 pixels. The Figure 4 shows an image of the 
Calibration and Test Site captured by the second camera: it is distorted by a jitter of about 2 to 4 pixels. The vertical 
lines are plumb lines and should be straight lines on the image. 
  
Figure 4: An example of an image distorted by a jitter of about 3 to 4 pixels 
Considering that a image distortion in the form of a jitter larger than one ore two pixels will make this images unusable 
for a high quality cartographic work, it was decided to test a new configuration using two frame grabbers. It was a wise 
decision since this time we acquired distortion free images in analogue mode. However, the system was not totally 
stable, some times an unexpected, mild jitter was present (maximum 1 pixel). The system was now working also in 
digital mode. The digital images were good (better than in the analogue mode) but sometimes distortions were present 
and some “drop out” occurred. 
With the help of the Electrical Engineering Department, a new set up was installed whereby the synchronization signals 
(vertical and horizontal) are now generated by the first Matrox frame grabber card. This first card works as the “master” 
card, controlling the synchronization of both cameras. The digital images delivered now by the system, using the 
described configuration, are distortion free, high quality images. 
3 EXPERIMENTATION IN REAL FLIGHT CONDITIONS 
Before the test in real flight condition, a simulation was organized inside the laboratory. The camera mount and the 
mobile computer containing the frame grabbers were installed on a mobile platform. The camera mount was oriented 
horizontally, perpendicular to the wall where about 50 targets were installed and measured using micro geodesy 
techniques. The power supply was provided by batteries, thus simulating real-time conditions. 
The test area for the real flight test was chosen over the campus of the Laval University in Quebec City. A number of 
control points were selected, most of them being natural details of the terrain. With the help of our colleagues from the 
ace the old Department of Geomatic Sciences, about 24 points were measured using the differential GPS technique. These points 
d to should have a precision of better then 10 cm while the image captured during the test flight should have a pixel size of 
Is. It has about 40 by 40 cm. A flight plan was prepared for a small block of 3 lines with an along track overlap of about 75 % to 
if it works 80 %. 
The first test flight took place in October 1999 following the prepared flight plan. The hardware and software 
ratory but Y C x 
© y components for the direct, near real time image data download from the airplane were not operational at the moment of 
  
International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part Bl. Amsterdam 2000. 269