Multi-class segmentation of navel orange surface defects based on improved DeepLabv3+

Published: 20 February 2024
Abstract Views: 405
PDF: 38
Publisher's note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.


A multi-class segmentation model based on improved DeepLabv3+ is proposed to detect navel orange surface defects. This model aims to address the problems of the current mainstream semantic segmentation network, including rough edge segmentation of navel orange defects, poor accuracy of small target defect segmentation, and insufficient deep-level semantic extraction of defects, which will result in the loss of feature information. In order to improve semantic segmentation performance, the Coordinate Attention Mechanism is integrated into the DeepLabv3+ network. Additionally, the deformable empty convolution of the Atrous Spatial Pyramid Pooling structure replaces the dilated convolution, improving the network’s ability to fit and target irregular defects and shape changes. Furthermore, to achieve multi-scale feature fusion and enhance feature space and semantic information, a Bi-feature pyramid network-based feature fusion branch is added at the DeepLabv3+ encoder side. The experimental findings demonstrate that the improved DeepLabv3+ model improves the extraction capability of navel orange defect features and has better segmentation performance. On the navel orange surface defect dataset, the improved model’s average intersection ratio and average pixel intersection ratio accuracies are 77.32% and 86.38%, respectively, which are 3.81% and 5.29% higher than the original DeepLabv3+ network.



PlumX Metrics


Download data is not yet available.


Bhargava, A., Bansal, A. Automatic detection and grading of multiple fruits by machine learning. 2020. Food Anal. Methods 13:751-61.
Chen, L.C., Zhu, Y., Papandreou, G., Schroff, F., Adam, H. 2018. Encoder-decoder with atrous separable convolution for semantic image segmentation. In Proceedings of the European conference on computer vision (ECCV). pp. 801-18.
Fan, S., Li, J., Zhang, Y., Tian, X., Wang, Q., He, X., Zhang, C., Huang, W. 2020. Online detection of defective apples using computer vision system combined with deep learning methods. J. Food Eng. 286:110102.
Hou, Q., Zhou, D., Feng, J. 2021. Coordinate attention for efficient mobile network design. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 13713-22.
Hu, J., Shen, L., Sun, G. 2018. Squeeze-and-excitation networks. In Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 7132-41.
Jin, L.I., Renyong, Z.H.A.O., Boxue, D.U., Liucheng, H., Kai, B. 2021. Research progress of nondestructive detection methods for defects of electrical epoxy insulators. Transact. Chin. Electrotechn. Soc. 36:4598-607.
Li J.B., Huang W.Q., Zhao C.J. 2015. Machine vision technology for detecting the external defects of fruits - A review. Image. Sci. J. 63:241-51.
Liang, X., Jia, X., Huang, W., He, X., Li, L., Fan, S., Li, J., Zhao, C., Zhang, C. 2022. Real-time grading of defect apples using semantic segmentation combination with a pruned YOLO V4 Network. Foods. 11:3150.
Lin, T.Y., Dollár, P., Girshick, R., He, K., Hariharan, B., Belongie, S. 2017. Feature pyramid networks for object detection. Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 2117-25.
Nithya, R., Santhi, B., Manikandan, R., Rahimi, M., Gandomi, A.H. 2022. Computer vision system for mango fruit defect detection using deep convolutional neural network. Foods. 11:3483.
Raman, S., Chougule, A., Chamola, V. 2022. A low power consumption mobile based IoT framework for real-time classification and segmentation for apple disease. Microprocess. Microsyst. 94:104656.
Ren, H.-E., Bai, J.-Y. 2013. Color grading based on dynamic clustering in lab color space. Jisuanji Gongcheng/Comput. Eng. 39.
Rong, D., Rao, X., Ying, Y. 2017. Computer vision detection of surface defect on oranges by means of a sliding comparison window local segmentation algorithm. Comput. Electron. Agr. 137:59-68.
Ronneberger, O., Fischer, P., Brox, T. 2015. U-net: convolutional networks for biomedical image segmentation. Medical Image Computing and Computer-Assisted Intervention 2015: 18th International Conference, Munich, Germany, October 5-9, 2015, Proceedings, Part III 18. Springer International Publishing. pp. 234-41
Soltani Firouz, M., Sardari, H. 2022. Defect detection in fruit and vegetables by using machine vision systems and image processing. Food Eng. Rev. 14:353-79.
Strudel, R., Garcia, R., Laptev, I., Schmid, C. 2021. Segmenter: transformer for semantic segmentation. In Proceedings of the IEEE/CVF international conference on computer vision. pp. 7262-72.
Sun, X., Li, G., Xu, S. 2020. Fastidious attention network for navel orange segmentation. arXiv preprint arXiv:2003.11734.
Tian, K., Zeng, J., Song, T., Li, Z., Evans, A. Li, J. 2022. Tomato leaf diseases recognition based on deep convolutional neural networks. J. Agric. Eng. Res. 54.
Unay, D. 2022. Deep learning-based automatic grading of bi-colored apples using multispectral images. Multimed. Tools Appl. 81:38237-52.
Woo, S., Park, J., Lee, J.Y., Kweon, I.S. 2018. Cbam: Convolutional block attention module. Proceedings of the European Conference on Computer Vision. pp. 3-19.
Xie, E., Wang, W., Yu, Z., Anandkumar, A., Alvarez, J.M. Luo, P. 2021. SegFormer: simple and efficient design for semantic segmentation with transformers. Adv. Neur. In. 34:12077-90.
Xie, X., Ge, S., Xie, M., Hu, F., Jiang, N., Cai, T., Li, B. 2018. Image matching algorithm of defects on navel orange surface based on compressed sensing. J. Amb. Intel. Hum. Comp. pp. 1-9.
Yang, G.L., Luo, L., Feng, Y.Q., Zhao, H.S. 2014. Research of navel orange defect and color detection based on machine vision. Appl. Mech. Mater. 513:3442-5.
Yao, J., Qi, J., Zhang, J., Shao, H., Yang, J., Li, X. 2021. A realtime detection algorithm for Kiwifruit defects based on YOLOv5. Electronics. 10:1711.
Yu, C., Gao, C., Wang, J., Yu, G., Shen, C. Sang, N. 2021. Bisenet v2: Bilateral network with guided aggregation for real-time semantic segmentation. Int. J. Comput. Vision. 129:3051-68.
Zhang, B., Huang, W., Gong, L., Li, J., Zhao, C., Liu, C., Huang, D. 2015. Computer vision detection of defective apples using automatic lightness correction and weighted RVM classifier. J. Food Eng. 146:143-51.
Zhao, H., Shi, J., Qi, X., Wang, X., Jia, J. 2017. Pyramid scene parsing network. Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 2881-90.
Zhou, H., Zhuang, Z., Liu, Y., Liu, Y., Zhang, X. 2020. Defect classification of green plums based on deep learning. Sensors. 20:6993.

How to Cite

Zhu, Y., Liu, S., Wu, X., Gao, L. and Xu, Y. (2024) “Multi-class segmentation of navel orange surface defects based on improved DeepLabv3+”, Journal of Agricultural Engineering, 55(2). doi: 10.4081/jae.2024.1564.