<title>Detection and clutter rejection in image sequences based on multivariate conditional probability</title>

Abstract

A method of detecting dim targets in highly-cluttered time-varying image sequences is presented, where reliable clutter rejection is achieved by calibrating the multivariate statistics of a small number of generic space-time filters. The targets have sufficiently low SCR that a track-before-detect method is required. For targets where there is little prior information on velocity, a large number of filters is generally required to achieve a high response relative to the background. In the method described here, instead of applying thresholds to individual filters, joint filter statistics are used to estimate conditional threshold exceedance probabilities. A smaller number of more generic filters are applied, which are not finely tuned to targets but which characterise aspects of both targets and clutter. Potential targets are cued based on a non-parametric estimate of the probability of occurrence of similar clutter. Constant fal se alarm rates are inherent in the method. The method is demonstrated on examples of real forward-looking imagery of the sea surface, where glint is a source of strong clutter. Dim targets are distinguished from clutter by using the joint statistics of three variables: a constant-intensity blob filter, a filter tuned to sea glint flashes, and the vertical image co-ordinate.

An new algorithm of infrared moving small targets recognition

“…We also notice that, there are same features of the target regions in the image after filtering whichever kind filter is chosen, which have a high intensity. So G.H.Waston presented a method based on multivariate conditional probability [7] , but as he pointed out that the method was hard to be realized for the problem is often an ill-posed.…”

Adaptive Automatic Target Recognition with SVM Boosting for Outlier Detection

“…In [12] a target recognition approach was proposed in which the multivariate statistics of space-time structure was used to characterise spatially and temporally highly structured clutter, such as sea-glint and atmospheric scintillation. Targets were then recognised as unusual events.…”

“…This method is compared to another TBD technique [12] in which targets are distinguished from the clutter by the analysis of the joint statistics of simple events such as glint flashes and regions of persistent brightness.…”

Automated object extraction from remote sensor image based on adaptive thresholding technique

“…Summarily, those algorithms can be divided into two kinds. One kind of method is processing single images followed by association and tracking, which includes Matched Filtering (MFT), Dynamic Programming Algorithm (DPA) [3] , Multistage Hypothesis Testing (MAHT) [4] , and so on. This kind of method emphasizes the intensity features of the targets in a single image and their movement continuities in the image sequences.…”