we investigate the feasibility of using RGB images exclusively as input data instead of multi-spectral images. Encouragingly, the results show that the accuracy achieved in this scenario is nearly comparable to the multi-spectral approach. The experiments can be reproduced using the code made available at the following address: https://github.com/cimice15/Quasi_real-time_apple_defect_segmentation (accessed on 8 September 2023). The paper is organized as follows: Section 2 presents related works, Section 3 presents the database used in our experiments and the method we propose. Section 4 presents evaluation metrics and experimental setups. Finally Section 5 discusses results of the proposed m
Open resource ↗cimice15/Quasi_real-time_apple_defect_segmentation · lines:40-49Unverified paper record
Quasi Real-Time Apple Defect Segmentation Using Deep Learning.
Sensors (Basel, Switzerland) · 14 Sept 2023 · 10.3390/s23187893
Abstract
Defect segmentation of apples is an important task in the agriculture industry for quality control and food safety. In this paper, we propose a deep learning approach for the automated segmentation of apple defects using convolutional neural networks (CNNs) based on a U-shaped architecture with skip-connections only within the noise reduction block. An ad-hoc data synthesis technique has been designed to increase the number of samples and at the same time to reduce neural network overfitting. We evaluate our model on a dataset of multi-spectral apple images with pixel-wise annotations for several types of defects. In this paper, we show that our proposal outperforms in terms of segmentation accuracy general-purpose deep learning architectures commonly used for segmentation tasks. From the application point of view, we improve the previous methods for apple defect segmentation. A measure of the computational cost shows that our proposal can be employed in real-time (about 100 frame-per-second on GPU) and in quasi-real-time (about 7/8 frame-per-second on CPU) visual-based apple inspection. To further improve the applicability of the method, we investigate the potential of using only RGB images instead of multi-spectral images as input images. The results prove that the accuracy in this case is almost comparable with the multi-spectral case.
Plant phenotyping relevance
リンゴ果実の欠陥を画像から画素単位で抽出する深層学習手法を開発・評価しており、植物器官の状態（欠陥）取得が中心的な方法論的貢献である。
abstractwe propose a deep learning approach for the automated segmentation of apple defects using convolutional neural networks (CNNs)
abstractWe evaluate our model on a dataset of multi-spectral apple images with pixel-wise annotations for several types of defects.
abstractwe show that our proposal outperforms in terms of segmentation accuracy general-purpose deep learning architectures commonly used for segmentation tasks.
Code and data availability
The paper's authors publicly release the analysis code for their apple defect segmentation experiments via a GitHub repository, explicitly stated in the text. The apple image dataset itself is cited prior work (Kleynen et al.) and no separate dataset deposit by these authors is stated.
This is an automatically classified, unverified record. Curator approval is required before any resource enters the Catalog.