It is a challenging task to build efficient and robust model for aircraft detection. In our object recognition system, aircraft detection is a main task, which faces various problems, such as blur, occlusion, and shape variation and so on. Existing approaches always require a set of complex classification model and a large number of training samples, which is inefficient and costly. In order to deal with these problems, we employ location based informative features to reduce the complexity of training data. With the employment of location based informative features, simple classifiers will manifest high performance instead of complex classifier which requires more complicated strategy for training. Further, our system needs to update the model frequently which is similar to online learning method, in order to reducing computational complexity, a very sparse measurement matrix is applied to extract features from feature space. The construction of this sparse matrix is based on the theory of sparse representation and compressed sensing. From the experimental results, the detection rate and cost of our proposed method is better than other traditional method.
An approach is proposed to reduce the tracking jitter of the extended target in boost phase for plume tracker in a photoelectric acquisition, tracking, and pointing system. The characteristics of the vehicle imaging are analyzed and the causes of jitters are identified. The target moving direction and its principal axis are combined to calculate the optimal frontal direction. A contour smoothing method based on the chord-arc ratio filtering is introduced to obtain a preliminary extraction point with lower jitters. Then a fine tracking point extraction method based on the minimal inscribed circle of contour after filtering is presented. Experimental results confirm that the proposed method significantly improves the tracking precision and stability.
Automatic focusing (AF) is a key technology of measuring TV capturing the clear objective image in photoelectric
measurement system. It is viable to enhance the performance of measuring TV through focusing effectively and quickly.
In the process of maneuvering target tracking, the background and the feature of targets change from time to time, and
the reliability of AF is highly required. Firstly, conditions for starting AF need to be investigated. The relation between
degree of definition and edge acutance is proved by experiments. Combined with the sharpness value, it decides whether
to begin AF. Secondly, it needs focusing quickly and exactly after starting AF. The accuracy and efficiency of the
sharpness function is another key factor of AF. By comparing some favorable sharpness functions, normalized variance
and square-gradient functions are employed based on focus windows. Thirdly, the optimized mountain-climb searching
algorithm based on the defocusing extents and the adaptive searching step size is proposed. Experiments show the
algorithm proposed improves the speed and reliability of AF.