CS Paper Radar

Hierarchical Filter Band Selection for Multispectral Object Classification

Katja Kossira, Jürgen Seiler, André Kaup

eess.IV ·submitted 2026-09-20 14:27 UTC · v1
published

Abstract

Multispectral camera arrays capture image data in various spectral bands, enabling image acquisition beyond human perception. These systems are widely used in medical, agricultural, environmental, and remote sensing applications. However, not all recorded bands are needed for classification tasks, thus reducing them can lower hardware complexity and cost. The conditional filter band selection algorithm addresses this by selecting low-noise, non-redundant bands to minimize the number of filters and cameras. This paper improves the approach by introducing a second classification stage that estimates object material in addition to the object label. This information is merged by a decision-tree based band selection strategy. The proposed method achieves a 28.9% relative reduction in classification error on the SMM50 dataset compared to the state-of-the-art. Moreover, for the same classification accuracy, the required number of cameras is reduced from 7 to 4, demonstrating that the proposed approach improves performance while significantly lowering hardware requirements.

Comment
This paper was publsihed at IEEE International Conference on Image Processing (ICIP) 2026, Tampere, Finnland
DOI
10.1109/ICIP61757.2026.11630382

Closest papers

Nearest neighbours by meaning across the whole index.

Multispectral Household Plastic Classification for Recycling Using a Camera Array

eess.IV ·2026-08-23 ·Katja Kossira, Jürgen Seiler, André Kaup
publishedcode/data similarity 0.69
Abstract

Plastic pollution has become a persistent problem in natural ecosystems. Driven by insufficient waste management, limited recycling efficiency, and high costs for recycling, plastic waste in the environment poses significant ecological and health risks. Recycling requires accurate identification of polymer types, but existing optical sorting systems often struggle to distinguish common household plastics. In this work, we present a novel classification approach based on a multispectral imaging system consisting of nine cameras equipped with near-infrared bandpass filters. The system is designed to discriminate the seven most common household plastics. From the resulting multispectral images, we extract the spectral fingerprints and derive features such as intensity differences between specific wavelength pairs and their slopes, as well as false-color image representations. A dedicated preprocessing pipeline aligns and normalizes the data before classification. We recorded a multispectral household plastic database (https://github.com/FAU-LMS/MHPM) and trained four different classifiers Gradient Boosting, Extreme Gradient Boosting, Light Gradient Boosting Machine, and CatBoost. The best-performing model achieves a classification accuracy of 86.7%. The computational runtime is 2.603 μs per pixel, enabling efficient processing of high-resolution images. The entire setup is built from off-the-shelf hardware components, which makes replication straightforward and allows direct integration into industrial sorting pipelines.

0.56

Band-Selection Stability and Semantic Segmentation Performance: A Study on Hyperspectral City

cs.CV ·2026-09-25 ·Jiarong Li, Imad Ali Shah, Enda Ward, Martin Glavin, Edward Jones +1 more
published similarity 0.58
Abstract · Accepted for IEEE WHISPERS 2026

Resource constraints make high-dimensional hyperspectral imaging challenging in autonomous perception, motivating the use of band selection methods. However, the sensitivity of band-selection methods to sampled data and their relationship to semantic segmentation models (SSMs) remain underexplored. This study evaluates six band selection methods on ten independently sampled, class-balanced region-of-interest (ROI) sets, yielding 60 top-25 band subsets from the Hyperspectral City V2 (128 bands: 450-950nm) dataset. Top-$K$ bands ($K\in\{3,5, ... 13\}$) from the first three ROI sets are evaluated with three SSMs against the corresponding 128-band baseline. Experiments show that intra-method stability is method-dependent: Sim-LP shows the highest stability (pairwise Jaccard similarity) and, together with JMIM+CSNR, yields the best segmentation results. Top-$K$ based SSMs remain competitive with baselines, with gains of up to 2.01 mIoU and 1.72 mF1 points, and 18-22x faster CPU inference for $K=9$. However, performance does not improve monotonically with $K$, and stability shows no consistent association with SSM performance. These findings suggest that intra-method stability is informative but an unreliable indicator of downstream segmentation performance, highlighting the need to evaluate band-selection methods across repeated samples, subset sizes, and SSMs.

0.37

Spectral Haar Binarizer: Compact Boolean Contrast Representations for Hyperspectral Classification

cs.LG ·2026-10-30 ·Michał Romaszewski, Przemysław Głomb
similarity 0.57
Abstract

Tsetlin Machines offer an interpretable, hardware-friendly alternative to conventional machine learning models, but applying them to continuous hyperspectral measurements requires converting spectral signals into Boolean representations. Standard binarisation methods typically discretise each spectral band independently, potentially discarding important relationships between neighbouring wavelengths. In this paper, we introduce the Spectral Haar Binarizer (SHB), a contrast-based spectral representation method that constructs Boolean predicates from differences between selected spectral intervals. Inspired by Haar-like contrast features, the method uses an AdaBoost-style selection procedure to obtain a compact set of discriminative spectral predicates. We evaluate SHB on five heterogeneous hyperspectral datasets covering forensic blood detection, agricultural and urban remote sensing, and rainforest tree-species classification. Across binary and multiclass tasks, SHB consistently outperforms standard per-band binarisation strategies and provides the strongest or near-strongest results across classifiers. In binary tasks, TM-SHB achieves the best average performance, while RF-SHB is strongest in multiclass settings. Bayesian and rank-based analyses confirm that SHB's advantage is robust across datasets and not driven by isolated cases. The results show that representing continuous hyperspectral measurements is a central processing step for efficient classification. That contrast-based Boolean features provide an effective bridge between information-rich sensor measurements and compact, logic-based processing.

0.26

Dimensionality Reduction for Hyperspectral Image Classification

cs.CV eess.SP ·2026-09-09 ·Mohamed Cherifi, Ammar Mesloub, Mohammed Nabil El Korso, Tayeb Touhami, Abdennour Hacine Gharbi
published similarity 0.53
Abstract · Published in 2024 8th International Conference on Image and Signal Processing and their Applications (ISPA), Biskra, Algeria

This paper addresses the issue of supervised classification in the context of hyperspectral satellite images. It deals with two fundamental aspects: dimensionality reduction of data and the selection of appropriate supervised classification techniques. Firstly, we delve into dimensionality reduction, a critical step in simplifying the management of hyperspectral data. The reduction aims to decrease complexity in terms of memory and computing time. We examine two commonly used methods: Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA). Subsequently, we explore the selection of the most suitable supervised classification algorithms for hyperspectral images. We compare the performance of three methods: K-Nearest Neighbors (KNN), Support Vector Machines (SVM), and Random Forest (RF) using real hyperspectral data. The results highlight that the combination of PCA and RF yields the highest overall accuracy and Kappa coefficient.

0.68

Vegetation Health Classification for Low-Data and Edge Hardware Deployment

cs.CV ·2026-11-14 ·Yifei Cheng, Haoyu Li
hot direction similarity 0.53
Abstract

Unmanned aerial vehicles (UAVs) equipped with cameras are popular, viable tools for vegetation and environmental health monitoring. However, existing image classification methods are incompatible with UAV implementation for independent practitioners in remote, resource-constrained environments due to large dataset requirements, GPU hardware, inflexible retraining, and low interpretability. This paper presents a lightweight, interpretable vegetation classification algorithm operating with as few as five reference images per class, without GPUs or heavy machine learning frameworks. It uses a 2D hue-saturation histogram feature representation and Earth Mover's Distance or Quadratic Form Distance, paired with a log-likelihood probability score, to classify images into three vegetation health categories. We evaluated six systems: three custom pipelines and three established models across low-data scenarios ($n = 5, 10, 25$) and a full-data regime of 2,172 drone images from 28 South American locations. Our custom pipelines achieve competitive Top-1 accuracy (81--85\%) under low-data constraints, maintaining sub-1ms inference times that enable real-time classification and subsequently tasks such as seeding and watering. The algorithms also support incremental retraining without full model reconstruction. We validate edge feasibility by deploying the QFD variant on an ESP32 microcontroller, achieving 82.8\% accuracy with 1.01\% heap memory utilization (45kb), demonstrating that effective vegetation classification is achievable on low-cost, low-power embedded hardware without cellular connectivity.

0.28

MWIR-4-Plastic: The Identification of Complex End-of-Life Industrial Plastic using Mid-wave Infrared Hyperspectral Imaging and Machine Learning

cs.CV cs.LG ·2026-08-28 ·Elias Arbash, Andréa de Lima Ribeiro, Filipa Simões, Ahmed Jamal Afifi, Aldino Rizaldy +7 more
publishedbenchmarkdatasetcode/data similarity 0.52
Abstract · 15 pages

The automated sorting of shredded black plastics from end-of-life (EOF) industrial waste presents a significant challenge in recycling facilities, primarily due to the limitations of current sensing and analytical approaches. Existing studies predominantly rely on single-point contact-based mid-infrared spectroscopy or laboratory hyperspectral imaging (HSI) setups, which fail to provide the spatially resolved analysis necessary for fast, bulk processing. Moreover, available datasets are laboratory-controlled and focus on intact rather than shredded plastics, hindering further recycling refinement. Black industrial plastics, in particular, are underrepresented, while most classification pipelines depend on manual region selection and rule-based spectral matching, neglecting spatial information and modern deep learning (DL) methods. To address these gaps, we introduce the first publicly available HSI dataset of shredded black plastics from EOF vehicle, comprising four industrial polymers across 13 co-registered RGB, VNIR, SWIR, and MWIR scenes and their segmentation pipeline. We developed a multi-modal spectral-spatial framework that integrates foreground isolation, pixel-wise classification, and object-level majority voting. By adapting advanced hyperspectral transformers from earth observation and incorporating chemometric band selection, we achieve accurate classification of complex black plastics. The study establishes the first comprehensive benchmark using nine processing methods, including chemometric, machine learning, and DL architectures. To ensure reproducibility, the complete dataset and methodologies are publicly released, establishing a benchmark for a hyperspectral object-analysis pipeline in industrial inspection.

0.31

Dimensionality reduction for AI based hyperspectral image classification based on XAI

cs.CV cs.AI ·2026-09-16 ·Vladimir Zeljković, Branka Stojanović, Harald Ganster, Aleksandar Nešković
similarity 0.50
Abstract · The paper has been presented at the Annual Workshop of the Austrian Association for Pattern Recognition (OAGM) 2023 - Patterns in One Health…

This research addresses the challenge of limited material recycling in wood recycling processes by leveraging artificial intelligence (AI)-based dimensionality reduction. Our study explores the application of convolutional neural networks (CNNs) in multi-channel hyperspectral imaging (HSI), extending beyond RGB channels to over 200 spectral channels. Dimensionality reduction within this context involves streamlining the feature space for AI system training and inference. Focusing on explainable AI (XAI) methods, this paper contributes to a broader research initiative, presenting a solution framework that enhances the sustainability and efficiency of wood recycling processes.

0.28

Seeing the Unseen: Camouflaged Object Detection Beyond the Visible Spectrum

cs.CV ·2026-08-31 ·Avi Gupta, Trasha Gupta
hot direction similarity 0.50
Abstract

Recent advances in camouflaged object detection (COD) have led to substantial progress in challenging low-visibility scenarios, with pioneering studies demonstrating notable success in localizing objects in camouflaged scenes. Despite these achievements, existing approaches predominantly rely on conventional three-channel RGB imagery, thereby constraining the available visual information to a limited spectral range. Multispectral images offer a wide range of information about a scene by capturing fine-grained spectral signatures. Hence, by leveraging multispectral images for COD, we introduce a novel approach to detect camouflaged objects from the corresponding multispectral inputs. In particular, we propose an end-to-end framework, \textbf{\textit{MSFormer}}, that takes a multispectral camouflaged image as input and predicts a binary mask for it. Additionally, we also provide empirical justification for integrating multispectral bands for this complex low-vision task. Our extensive experiments demonstrate the effectiveness of our method, which outperforms existing methods.

0.34