Research Of Knowledge-aided Hough Transform Algorithm

Due to the stealth technology, the advent of the cruise missile and the complicatedenvironment, the survival of warning radar faces severe test. Track before detection(TBD) method，by using backend radar signal processing method，can enhance the radarweak target detection performance. As Hough transform TBD(HT-TBD) technology issuitable for multi-object tracking and real-time application, it becomes a research focusin the international. Effective use of prior information of knowledge assistant system,can remove the influences of the target diversity and the complex environment to thealgorithm.This article mainly studied the improvement of HT-TBD after using the knowledgeassistant system. The content is organized as follows:1. As the detection performance of the2D Hough Transform TBD is poor, westudied the Modified Hough Transform algorithm(MHT)，compared to HT-TBD，theMHT-TBD has the better detection performance. Analyzed the advantages of theKA-HT-TBD.2. For the setting up of two criterion conditions of MHT, we put forward theParameter setting modes of criterion condition1and criterion condition2using thetarget’s information. The method is effective in reducing the false track of MHT-TBDand improving the probability of target detection3. As the clutter distribution of the actual environment tend to be homogeneous，we using the prior information of Clutter partition, puts forward an adaptive secondthreshold in MHT, then puts forward the selection method of the first threshold and theadjustment method of tracking strategy for MHT, take advantages of these methods caneffectively track moving targets in complex environment.4. As there always be some Isolated scatterers in the radar environment, we usingthe information of the strong discrete clutter points which GIS provides, puts forward asmoothing method to deal with the strong discrete clutter points in HT-TBD，themethod effectively eliminated the interference of strong discrete clutter points inHT-TBD.