Camera systems are often unsteady on platform of airborne, car borne and ship borne. Stabilization algorithm can
be used to eliminate impact of vibration. But image sequence after processing is different from original sequence. If
there is a moving target in camera field, feature points on the target must be indentified and made sure corresponding
relationship in processed sequence. To solve the problem that moving target features position and correspondence are
difficult to identify in image sequence after image stabilization processing, background updating difference moving target
detection algorithm based on motion analysis is proposed. It uses subsample mean and subsample variance and introduces
the concept of background gray probability to identify feature points of moving target in the steady image sequence. In
addition, to solve the problem of incomplete motion track of feature points caused by obstruction or weak target detection
algorithm, partial limit incomplete smooth track algorithm is proposed. It is used to identify correspondence of feature
points on the moving target, and to solve temporary occlusion of moving object. Experimental results show that moving
target features position and correspondence can be identified quickly through the two algorithms. Single-frame processing
speed can reach an average of 27 ms with DSP6416 processor. Image stabilization algorithm and the two algorithms can be
combined to realize real-time tracking based on image stabilization.
Object tracking technology combined with image stabilization is called tracking technology based on image
stabilization. Moving objects affect stabilization compensation in tracking algorithm based on image stabilization. The
middle value method, former background method and dynamic clustering method are not useful for rotate or non-rigid
objects. Fuzzy Clustering of feature points is proposed to solve these problems. First background and objectives
membership value of the pixel are calculated, and then pixels are classified to background and target categories
accurately by defining membership threshold of the background and targets. Experimental results show that fuzzy
clustering algorithm resolves the moving target interference problem in image sequence and realizes steady tracking of
object. It is proved to be more robust and less sensitive to the numbers of initial clustering.
An improved corner detection algorithm based on SUSAN principle is proposed. Because SUSAN operator is hard to
distinguish the corner from some special points on the digital image edges, a double template is constructed. It extracts
potential corners by SUSAN operator and then decides the accurate location of corners by a 5×5 template. Meanwhile,
an adaptive selection of gray threshold t is proposed on the basis of the local gray discreteness of pixel. The experiment
results show that the improved algorithm further raises the accuracy of corner detection and is more suitable for
application in digital image processing.
Image stabilization can be used in variety of situations including tracking system on an unstable platform. In
order to realize this aim, some questions need to be solved, including feature point matching, impact of moving
object, existence of abnormal value, how to make certain feature points on the target, and so on. In this article, a
new image stabilization algorithm for digital image tracking is proposed and resolved upon questions by it. The
experiment result shows that this method can realize object tracking based on image stabilization and prove its
significance in practical application.