Multiscale Superpixelwise Prophet Model for Noise-Robust Feature Extraction in Hyperspectral Images

Abstract

Despite of various approaches proposed to smooth the hyperspectral images (HSIs) before feature extraction, the efficacy is still affected by the noise, even using the corrected dataset with the noisy and water absorption bands discarded. In this study, a novel spectral-spatial feature mining framework, Multiscale Superpixelwise Prophet Model (MSPM), is proposed for noise-robust feature extraction and effective classification of the HSI. The prophet model is highly noise-robust for deeply digging into the complex structured features thus enlarging interclass diversity and improving intraclass similarity. First, the superpixelwise segmentation is produced from the first three principal components of an HSI to group pixels into regions with adaptively determined sizes and shapes. A multiscale prophet model is utilized to extract the multiscale informative trend components from the average spectrum of each superpixel. Taking the multiscale trend signal as the input feature, the HSI data are classified superpixelwisely, which is further refined by a majority vote based decision fusion. Comprehensive experiments on three publicly available datasets have fully validated the efficacy and robustness of our MSPM model when benchmarked with eleven state-of-the-art algorithms, including six spectral-spatial methods and five deep learning ones. Besides, MSPM also shows superiority under limited training samples, due to the combined strategies of superpixelwise fusion and multiscale fusion. Our model has provided a useful solution for noise-robust feature extraction as it achieves superior HSI classification even from the uncorrected dataset without prefiltering the water absorption and noisy bands.

An AIoT-Based Assistance System for Visually Impaired People

“…In the future, to further improve its detection accuracy with the help of sufficiently large data samples, several advanced models such as contrastive learning approach [52], self-attention enhanced deep residual network [53], time-series sequencing method [54], multiscale superpixelwise prophet model [55], and multistage stepwise discrimination with compressed MobileNet [56] will be investigated into the assistance system. Furthermore, additional functions, such as emotion recognition and fatigue detection, will be designed to enhance the overall life quality of visually impaired people.…”

A Fast Adaptive Binarization Method for QR Code Images Based on Dynamic Illumination Equalization

“…In the future, to achieve the integration of the proposed method into embedded devices, the algorithm will be implemented into the QR code decoder. Additionally, to further enhance the edge detection to improve the image recognition rate, several advanced methods in related fields, such as the end-to-end cross band 2D attention network [38], multiscale superpixelwise prophet model [39], self-attention enhanced deep residual network [40], and multistage stepwise discrimination with compressed MobileNet [41], will be investigated.…”

Omni-Domain Feature Extraction Method for Gait Recognition

“…The existing feature extraction backbone mainly uses a well-designed model to focus on feature extraction, such as [34][35][36][37][38][39][40], or weakly supervised/self-supervised methods for feature extraction, such as [41][42][43][44][45], or the use of some smart sensors for auxiliary feature extraction, such as [46][47][48][49]. Among them, using a unified model for feature extraction is currently the most commonly used method for vision-based gait recognition, and weakly supervised or even unsupervised learning is currently less used in the field of gait recognition, because it does not require data to be labeled.…”