Hyperspectral image is a three-dimensional data cube which describes spatial information and spectral information of the scene. The anomaly detection technique can detect the targets which have difference between the image and the background without priori information. Kernel independent component analysis(KICA) is a method of mapping hyperspectral data into the kernel space for feature extraction. In this paper, the hyperspectral image is subjected to abnormal information detection based on KICA. First, we calculate the kernel matrix K in order to map the data to high-dimensional space for whitening and dimension reduction processing. Then we utilize the FastICA algorithm to extract the core independent component (KIC). Finally, the extracted independent components with the most abnormal information are analyzed by RX operator, kernel RX operator and abundance quantization method. Comparing with the simulation result and the detected result by RX method, the representation shows the algorithm based on KICA has better detection performance.
It is very critical that make full use of the local information for infrared dim and small target tracking. In this paper, an effective and fast algorithm based on the context learning is proposed to track infrared dim moving target. Firstly, the principle of the spatio-temporal context learning algorithm is described and the tracking deviation is analyzed. Then, a correlation filter is utilized to get a rational context prior for the dim moving target, the advantage is that the prior considers the image intensity information between a target and its surround pixels. Furthermore, a Gaussian high-pass filter is introduced to extract an accurate spatial context, which has little influence caused by the cluttered background. At last, the tracking problem is posed by computing a confidence map which takes into account sufficient information of a dim target and its surround background. Since the proposed algorithm is realized using fast Fourier transform, it is easy to be real-time. The experiments on various clutter background sequences have validated the tracking capability of the proposed method. The experimental results show that the proposed method can provide a higher accuracy and speed than several classical algorithms, including the improved Template Matching algorithm, the Temporal-Spatial Fusion Filtering algorithm and the Moving Pipeline Filtering algorithm.