Near-Infrared Hyperspectral Target Tracking Based on Background Information and Spectral Position Prediction

Applied Sciences, Год журнала: 2025, Номер 15(8), С. 4275 - 4275

Опубликована: Апрель 12, 2025

In order to address the problems of in-plane rotation and fast motion during near-infrared (NIR) video target tracking, this study explores application capsule networks in NIR proposes a network method based on background information spectral position prediction. First, history frame extraction module is proposed. This performs matching images through average curve groundtruth value makes rough distinction between background. On basis, frames stored as pool for subsequent operations. The proposed routing combines traditional algorithm with information. Specifically, similarity feature space calculated, weight allocation mechanism dynamically adjusted. Thus, discriminative ability strengthened. Finally, prediction locates center search region next by fusing features adjacent current frame. effectively reduces computational complexity improves tracking stability. Experimental evaluations demonstrate that novel framework achieves superior performance compared methods, attaining 70.3% success rate 88.4% accuracy data. Meanwhile, visible spectrum (VIS) data analysis, architecture maintains competitive effectiveness 59.6% 78.8% precision.

Язык: Английский

---

Research on Camouflage Target Classification and Recognition Based on Mid Wave Infrared Hyperspectral Imaging

Remote Sensing, Год журнала: 2025, Номер 17(8), С. 1475 - 1475

Опубликована: Апрель 21, 2025

Mid-wave infrared (MWIR) hyperspectral imaging integrates MWIR technology with remote sensing, enabling the capture of radiative information that is difficult to obtain in visible spectrum, thus demonstrating significant value camouflage recognition and stealth design. However, there a notable lack open-source datasets effective classification methods this field. To address these challenges, study proposes dual-channel attention convolutional neural network (DACNet). First, we constructed four (GCL, SSCL, CW, LC) fill critical data gap. Second, issues spectral confusion between camouflaged targets backgrounds blurred spatial boundaries, DACNet employs independent branches extract deep spectral–spatial features while dynamically weighting through channel mechanisms, significantly enhancing target–background differentiation. Our experimental results demonstrate achieves an average accuracy (AA) 99.96%, 99.45%, 100%, 95.88%; overall (OA) 99.94%, 99.52%, 96.39%; Kappa coefficients 99.91%, 99.41%, 95.21% across datasets. The exhibit sharp edges minimal noise, outperforming five learning three machine approaches. Additional generalization experiments on public further validate DACNet’s superiority providing efficient novel approach for classification.

Язык: Английский

---

Comparative analysis of spectral variable selection methods for NIR-based multi-component detection of Xanthoceras sorbifolium Bunge seed kernels

Microchemical Journal, Год журнала: 2025, Номер unknown, С. 113128 - 113128

Опубликована: Фев. 1, 2025

Язык: Английский

---

Analysis of Reason to Global Warming Based on Heat Pattern Using Hyperspectral Imaging: Artificial Intelligence Application

Remote Sensing in Earth Systems Sciences, Год журнала: 2024, Номер unknown

Опубликована: Сен. 20, 2024

Язык: Английский

---