🤖 AI Summary
Conventional CNN-based person detection methods suffer significant performance degradation under pose variations or severe occlusion due to their inherent reliance on human shape priors. To address this limitation, we propose HitoMi-Cam—a lightweight, spectral reflectance–driven person detection method leveraging clothing’s intrinsic spectral signature rather than geometric shape. HitoMi-Cam pioneers shape-agnostic detection via spectral features and is the first such approach deployed and validated on real embedded hardware. By tightly integrating spectral imaging with edge-computing architecture, it achieves real-time inference at 23.2 fps on GPU-free, resource-constrained devices, with a mean average precision of 93.5%, outperforming state-of-the-art CNN baselines by up to 53.8 percentage points. Extensive experiments in simulated disaster search-and-rescue scenarios demonstrate its low false-positive rate and exceptional robustness to occlusion and pose variation. HitoMi-Cam thus serves as a complementary, reliable alternative to conventional vision-based detection systems in challenging operational environments.
📝 Abstract
While convolutional neural network (CNN)-based object detection is widely used, it exhibits a shape dependency that degrades performance for postures not included in the training data. Building upon our previous simulation study published in this journal, this study implements and evaluates the spectral-based approach on physical hardware to address this limitation. Specifically, this paper introduces HitoMi-Cam, a lightweight and shape-agnostic person detection method that uses the spectral reflectance properties of clothing. The author implemented the system on a resource-constrained edge device without a GPU to assess its practical viability. The results indicate that a processing speed of 23.2 frames per second (fps) (253 × 190 pixels) is achievable, suggesting that the method can be used for real-time applications. In a simulated search and rescue scenario where the performance of CNNs declines, HitoMi-Cam achieved an average precision (AP) of 93.5%, surpassing that of the compared CNN models (best AP of 53.8%). Throughout all evaluation scenarios, the occurrence of false positives remained minimal. This study positions the HitoMi-Cam method not as a replacement for CNN-based detectors but as a complementary tool under specific conditions. The results indicate that spectral-based person detection can be a viable option for real-time operation on edge devices in real-world environments where shapes are unpredictable, such as disaster rescue.