|
In airborne earth observation system, due to the movement of the flight and the vibration of the payload, the quality of
the picture captured will degrade severely because there exists a relative motion between the object and the imaging
system. With the development of high-performance, high-resolution CCD / CMOS technique and the further
optimization of imaging lens, the platform vibration becomes an important constraint in the high-resolution imaging, and
thus image stabilization for imaging systems is indispensable. The first step in image stabilization is to obtain the
vibration path by means of detecting the direction and length of the shake, and then to compensate for the shake using
some methods to eliminate the impact of vibration. In order to satisfy the need of fast displacement detection in the
image stabilization of airborne remote sensing system, we perform an auto displacement detection system based on joint
optical correlator in this paper. First we employ a vibration platform to simulate the scene with vibration in airborne
remote sensing system, and a high-speed camera system is set up to grab the target to form a video sequences with no
blur. Then two adjacent video frames will be input into a spatial light modulator parallelly in space, behind which a
Fourier lens is fixed to capture the spectrum of the object by a high speed camera. And after a high-pass filter, this
spectrum will be input into a spatial light modulator behind which fix another Fourier lens again, and another high speed
camera will capture the output image indicating the correlation peak position of the adjacent two images, from which
corresponding displacement vector can be calculated. Experiments show that this system works stably at a rate as high as
50 frames per second, with detection accuracy up to 0.1 pixel by centroid detection algorithm.
|
