Advances in Hyperspectral Image Processing Techniques

19Class Feature‐Weighted Hyperspectral Image Classification

Shengwei Zhong¹, Jiaojiao Li², Xiaodi Shang³, Shuhan Chen⁴, and Chein-I Chang⁵

¹School of Computer Science and Engineering, Nanjing University of Science and Technology, Nanjing, China

²State Key Laboratory of Integrated Services Networks, School of Telecommunications Engineering, Xidian University, Xian, China

³Center for Hyperspectral Imaging in Remote Sensing (CHIRS), School of Information and Technology, Dalian Maritime University, Dalian, China

⁴Department of Electrical Engineering, Zhejiang University, Hangzhou, China

⁵Remote Sensing Signal and Image Processing Laboratory, Department of Computer Science and Electrical Engineering, University of Marylandw Baltimore County, Baltimore, MD, USA

This chapter studies a recently developed class feature (CF)‐weighted hyperspectral image classification (CFW‐HSIC), which extracts features from classes that can be used to weight significance of each of classes for classification. Specifically, it can be used to address imbalanced class and background issues. Two types of class features (CFs) are of particular interest, intra‐class features (Intra‐CFs) and inter‐class features (Inter‐CFs) to capture class intra‐pixel and inter‐pixel information, respectively. These features are then used to calculate the CF probability for each of classes as its class significance for classification. Such obtained CF‐calculated probabilities not only automatically allocate an appropriate ...