Slowly Moving Target Detection Using t-SNE and Support Vector Machine

In this paper, a method using fractional signatures for small target detection is proposed based on fusion of features extracted from both the time-frequency domain and fractional domain by using principal component analysis (PCA) to get the key characteristics for redundancy reduction. The process of reducing feature dimensions is visualized by the t-distributed stochastic neighbor embedding (t-SNE) network, also the simulation based on real dataset offers better performance in small target detection under sea clutter environment.