Underwater Plankton Image Enhancement Method Based on Local Histogram Equalization
Abstract
Classification and identification of marine plankton are important for marine ecological environment monitoring. Due to the tiny size of individual plankton and the weak light intensity in the marine environment, it is challenging for the imaging system to capture a clear image of plankton in the flowing water. Sometimes causing weak brightness and noise interference in the collected images. To this end, we propose an underwater image enhancement method based on local histogram equalization to remove noise while improving image clarity effectively. The proposed method improves the brightness and contrast in HSI color space, and preserves the color information (hue and saturation) in H and S channels. Experiments were carried out on the underwater plankton images. The enhancement performance of the proposed method was evaluated by commonly used metrics and compared with the well-known image enhancement methods, including a deep learning-based method. The experimental results showed that the proposed method was superior to the other methods in noise reduction, structure similarity, and perceptual quality. This work presents a simple and effective method for underwater plankton image enhancement. The proposed enhancement method could provide high-quality images for implementing automatic plankton classification.
BibTeX
@inproceedings{Yang2024Underwater,
author = {Yang, Jin and Fan, Wenqiang and Fu, Longwen and Xia, Chunlei},
booktitle = {2024 IEEE 5th International Conference on Pattern Recognition and Machine Learning (PRML)},
title = {Underwater Plankton Image Enhancement Method Based on Local Histogram Equalization},
year = {2024},
volume = {},
number = {},
pages = {342-346},
abstract = {Classification and identification of marine plankton are important for marine ecological environment monitoring. Due to the tiny size of individual plankton and the weak light intensity in the marine environment, it is challenging for the imaging system to capture a clear image of plankton in the flowing water. Sometimes causing weak brightness and noise interference in the collected images. To this end, we propose an underwater image enhancement method based on local histogram equalization to remove noise while improving image clarity effectively. The proposed method improves the brightness and contrast in HSI color space, and preserves the color information (hue and saturation) in H and S channels. Experiments were carried out on the underwater plankton images. The enhancement performance of the proposed method was evaluated by commonly used metrics and compared with the well-known image enhancement methods, including a deep learning-based method. The experimental results showed that the proposed method was superior to the other methods in noise reduction, structure similarity, and perceptual quality. This work presents a simple and effective method for underwater plankton image enhancement. The proposed enhancement method could provide high-quality images for implementing automatic plankton classification.},
keywords = {Measurement;Histograms;Image color analysis;Plankton;Brightness;Noise reduction;Pattern recognition;Colored noise;Image enhancement;Monitoring;Plankton Image;Histogram Equalization;Underwater Image;Low-light enhancement},
doi = {10.1109/PRML62565.2024.10779784},
issn = {},
month = {July},
}