Early detection of apple bruises using spectral-spatial enhanced 3D CNN and region-based hyperspectral analysis.
Uložené v:
| Názov: | Early detection of apple bruises using spectral-spatial enhanced 3D CNN and region-based hyperspectral analysis. |
|---|---|
| Autori: | Liu C; Wuhan Polytechnic University, Wuhan 430023, China., Wu X; Wuhan Polytechnic University, Wuhan 430023, China., Yang W; Northwest A&F University, Yangling712100, China. Electronic address: ywq19960313@outlook.com., Zeng S; Wuhan Polytechnic University, Wuhan 430023, China., Wang J; Wuhan University, Wuhan 430072, China. |
| Zdroj: | Food research international (Ottawa, Ont.) [Food Res Int] 2025 Dec; Vol. 222 (Pt 1), pp. 117630. Date of Electronic Publication: 2025 Sep 27. |
| Spôsob vydávania: | Journal Article |
| Jazyk: | English |
| Informácie o časopise: | Publisher: Published on behalf of the Canadian Institute of Food Science and Technology by Elsevier Applied Science Country of Publication: Canada NLM ID: 9210143 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 1873-7145 (Electronic) Linking ISSN: 09639969 NLM ISO Abbreviation: Food Res Int Subsets: MEDLINE |
| Imprint Name(s): | Original Publication: Ottawa, Ontario, Canada : Published on behalf of the Canadian Institute of Food Science and Technology by Elsevier Applied Science, c1992- |
| Výrazy zo slovníka MeSH: | Malus* , Fruit* , Hyperspectral Imaging*/methods , Neural Networks, Computer*, Deep Learning ; Imaging, Three-Dimensional/methods |
| Abstrakt: | Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Hyperspectral imaging (HSI) has revolutionized the non-destructive detection of fruit defects by integrating spectral and spatial information. However, detecting early-stage minor bruises in apples remains challenging due to weak hyperspectral signals, high data dimensionality, and low computational efficiency. To address these issues, this study proposes a novel hyperspectral apple bruise detection network, termed 3D-HDI. The model constructs a backbone network using multiple 3D convolutions and integrates a feature enhancement module with a path aggregation network to amplify spectral signals and improve damage differentiation. Furthermore, it replaces pixel-by-pixel classification with a region-based detection head, significantly enhancing computational efficiency and accuracy. Experimental results demonstrate that the proposed model achieves a higher recognition rate (96.25%) while maintaining comparable detection efficiency, outperforming traditional classification networks such as 3D-EfficientNet, 3D-MobileNet, and 3D-AlexNet. This research advances the application of HSI and deep learning in fruit quality assessment, providing a robust solution for early-stage bruises detection. (Copyright © 2025 Elsevier Ltd. All rights reserved.) |
| Contributed Indexing: | Keywords: 3D convolution; Deep learning; Feature enhancement; Hyperspectral imaging; Non-destructive detection |
| Entry Date(s): | Date Created: 20251121 Date Completed: 20251121 Latest Revision: 20251121 |
| Update Code: | 20251121 |
| DOI: | 10.1016/j.foodres.2025.117630 |
| PMID: | 41267243 |
| Databáza: | MEDLINE |
Full Text Finder 
Nájsť tento článok vo Web of Science | Abstrakt: | Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.<br />Hyperspectral imaging (HSI) has revolutionized the non-destructive detection of fruit defects by integrating spectral and spatial information. However, detecting early-stage minor bruises in apples remains challenging due to weak hyperspectral signals, high data dimensionality, and low computational efficiency. To address these issues, this study proposes a novel hyperspectral apple bruise detection network, termed 3D-HDI. The model constructs a backbone network using multiple 3D convolutions and integrates a feature enhancement module with a path aggregation network to amplify spectral signals and improve damage differentiation. Furthermore, it replaces pixel-by-pixel classification with a region-based detection head, significantly enhancing computational efficiency and accuracy. Experimental results demonstrate that the proposed model achieves a higher recognition rate (96.25%) while maintaining comparable detection efficiency, outperforming traditional classification networks such as 3D-EfficientNet, 3D-MobileNet, and 3D-AlexNet. This research advances the application of HSI and deep learning in fruit quality assessment, providing a robust solution for early-stage bruises detection.<br /> (Copyright © 2025 Elsevier Ltd. All rights reserved.) |
|---|---|
| ISSN: | 1873-7145 |
| DOI: | 10.1016/j.foodres.2025.117630 |