Author(s)

Lu Li^#, Wei Fang^#, YiGe Shang, Zhi Zhang, MingFu Li, DaoYi Chen^*

Affiliation(s)

Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen 518000, Guangdong, China.

Corresponding Author

DaoYi Chen

ABSTRACT

Underwater barnacle cleaning robots require a semantic segmentation algorithm deployable onboard to effectively segment barnacles, obstacles, and background on ship hulls. This study independently constructed a dedicated ship hull semantic segmentation dataset and implemented an improved BiSeNet network. The approach integrates features from both a spatial path and a context path. Building upon the cross-entropy loss function, class weights were strategically assigned during training according to the distribution of different target categories within the dataset. This method achieved significantly high Intersection over Union (IOU) and F1 scores.

KEYWORDS

Spatial features; Contextual features; Semantic segmentation; Ship hull dataset;  Weighted loss function

CITE THIS PAPER

Lu Li, Wei Fang, YiGe Shang, Zhi Zhang, MingFu Li, DaoYi Chen. Semantic segmentation of ship hulls for underwater cleaning robots. Journal of Computer Science and Electrical Engineering. 2025, 7(6): 49-54. DOI: https://doi.org/10.61784/jcsee3087.

REFERENCES

[1] Trif László, Abdul Shaban, Judit Telegdi. Electrochemical and surface analytical techniques applied to microbiologically influenced corrosion investigation. Corrosion Reviews, 2018, 36(4): 349-363.

[2] Lander Karl. Proactive Cleaning, Maintenance and Inspection of Vessel Hulls via Autonomous Robots. Paper presented at the AMPP Annual Conference + Expo, Denver, Colorado, USA, 2023.

[3] Huang Q, Huang J. Comprehensive review of edge and contour detection: From traditional methods to recent advances. Neural Computing and Applications, 2025, 37(4): 2175-2209.

[4] S Kim, K Jo. BiSeNet with Depthwise Attention Spatial Path for Semantic Segmentation. 2022 International Workshop on Intelligent Systems (IWIS), Ulsan, Korea, Republic of. 2022, 1-4. DOI: 10.1109/IWIS56333.2022.9920717.

[5] N Das, S A Begum. An Empirical Study of Loss Functions for Aspect Category Detection in Imbalanced Data Scenario. 2025 10th International Conference on Signal Processing and Communication (ICSC), Noida, India. 2025, 247-252. DOI: 10.1109/ICSC64553.2025.10968230.

[6] G R S N Kumar, R S Sankuri, S P K Karri. Multi Scale aided Deep Learning model for High F1-score classification of fundus images based Diabetic Retinopathy and Glaucoma. 2023 International Conference on Computer, Electronics & Electrical Engineering & their Applications (IC2E3), Srinagar Garhwal, India. 2023, 1-6. DOI: 10.1109/IC2E357697.2023.10262434.

[7] Y Chen, X Li, Z He, et al. IoU_MDA: An Occluded Object Detection Algorithm Based on Fuzzy Sample Anchor Box IoU Matching Degree Deviation Aware. IEEE Access, 2024(12): 47630-47645. DOI: 10.1109/ACCESS.2024.3375109.

[8] Hennigh O, Narasimhan S, Nabian M A, et al. NVIDIA SimNet^TM: An AI-accelerated multi-physics simulation framework. International conference on computational science. Cham: Springer International Publishing. 2021, 447-461.

[9] Ferreira I O, Andrade L C, Teixeira V G, et al. State of art of bathymetric surveys. Boletim de Ciências Geodésicas, 2022, 28(01): e2022002.

[10] Wang X, Lin M, Li J, et al. Ultrasonic guided wave imaging with deep learning: Applications in corrosion mapping. Mechanical Systems and Signal Processing, 2022(169): 108761.