Advances in Hyperspectral Image Processing Techniques

3Efficient Hardware Implementation for Hyperspectral Anomaly and Target Detection

Jie Lei, Weiying Xie, Jiaojiao Li, Keyan Wang, Kai Liu and Yunsong Li

State Key Laboratory of Integrated Services Networks, Xidian University, Xi’an, China

3.1 Introduction

Hyperspectral images (HSIs) provide a wealth of spectral information [1] that is provided by the visible, near‐infrared, and shortwave infrared bands in remotely sensed scenes [2]. These spectra are divided into hundreds or even thousands of narrow contiguous spectral bands to capture more accurate measurements of the components of each pixel [3, 4]. Hyperspectral imaging has been widely applied to fields such as target detection [5, 6], anomaly detection [7], change detection [8], image classification [9, 10], band selection [11, 12], and hyperspectral unmixing [13, 14] due to its high spectral resolution. Among these applications, hyperspectral anomaly detection is of particular interest due to its capability of uncovering many unknown material substances [15, 16].

Real‐time target detection aiming to find timely interesting targets has drawn much attention because of its significance in military and civilian applications [17, 18]. Although the presence of targets in HSI provides critical information in data analysis, it is rather rare and may be difficult to find because many of the HSI‐detected targets are relatively small and weak. Consequently, real‐time target detection in HSI presents a great challenge and has become ...