Spatial Anomaly Detection in Hyperspectral Imaging Using Optical Neural Networks

Hyperspectral imaging (HSI) is a widely used technology, yet hard to implement in real-time anomaly detection due to its extensive data flow volume. An autoencoder structured hybrid optical–electrical neural network method is proposed in this work that realizes feature exaction and anomaly detection during the hyperspectral data acquisition process to address such issues. In the proposed method, a digital micromirror device functions as the core optical processor to extract low-level features from the hyperspectral data flow. Weight binarization and a conditional subnetwork are utilized to suit optical computation. Pretraining by artificial data is implemented to ease the training data burden. Case studies on defect detection and foreign object detection have demonstrated that the proposed method can significantly reduce the sampling time by orders of magnitude without loss of detection accuracy.

Hyperspectral imaging (HSI), also called imaging spectroscopy, is a hybrid technology of spectroscopy and imaging, obtaining two-dimensional images across a wide range of electromagnetic spectrums with high spectral resolution. Classified by dispersive devices, conventional HSI systems involve monochromator HSI, optical bandpass filter HSI, and single-shot imager HSI. Due to the high resolution of the spectral channel (bands), HSI is apt at classifying diverse objects of interest, resulting in its wide application in many areas, including military, medical, food safety control, agriculture, etc.