Article
An Embeddable Algorithm for Automatic Garbage Detection
Based on Complex Marine Environment
Hongjie Deng, Daji Ergu *, Fangyao Liu, Bo Ma and Ying Cai
Citation: Deng, H.; Ergu, D.; Liu, F.;
Ma, B.; Cai, Y. An Embeddable
Algorithm for Automatic Garbage
Detection Based on Complex Marine
Environment. Sensors 2021, 21, 6391.
https://doi.org/10.3390/s21196391
Academic Editor: Nikolaos Doulamis
Received: 11 September 2021
Accepted: 21 September 2021
Published: 24 September 2021
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2021 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Key Laboratory of Electronic and Information Engineering, Southwest Minzu University,
State Ethnic Affairs Commission, Chengdu 610041, China; denghj1221@163.com (H.D.); FLIU028@163.com (F.L.);
martbox@163.com (B.M.); 21500121@swun.edu.cn (Y.C.)
* Correspondence: ergudaji@163.com
Abstract:
With the continuous development of artificial intelligence, embedding object detection
algorithms into autonomous underwater detectors for marine garbage cleanup has become an
emerging application area. Considering the complexity of the marine environment and the low
resolution of the images taken by underwater detectors, this paper proposes an improved algorithm
based on Mask R-CNN, with the aim of achieving high accuracy marine garbage detection and
instance segmentation. First, the idea of dilated convolution is introduced in the Feature Pyramid
Network to enhance feature extraction ability for small objects. Secondly, the spatial-channel attention
mechanism is used to make features learn adaptively. It can effectively focus attention on detection
objects. Third, the re-scoring branch is added to improve the accuracy of instance segmentation by
scoring the predicted masks based on the method of Generalized Intersection over Union. Finally,
we train the proposed algorithm in this paper on the Transcan dataset, evaluating its effectiveness
by various metrics and comparing it with existing algorithms. The experimental results show that
compared to the baseline provided by the Transcan dataset, the algorithm in this paper improves
the mAP indexes on the two tasks of garbage detection and instance segmentation by 9.6 and 5.0,
respectively, which significantly improves the algorithm performance. Thus, it can be better applied
in the marine environment and achieve high precision object detection and instance segmentation.
Keywords:
deep learning; object detection; instance segmentation; marine ecology; the attentional
mechanism; dilated convolution
1. Introduction
The marine ecosystem is the essential condition for the survival and development of
marine organisms. Therefore, when the amount of external environmental changes exceeds
the tolerance limit of the biological community, it will directly affect the virtuous cycle of
the ecosystem, thus destroying the ecosystem.
In human development, manufactured or processed solid waste is inevitably discarded
into the ocean, and this category of waste is called marine litter. To some extent, it affects
the marine landscape, threatens the safety of shipping routes, and impacts the health of the
marine ecosystem, which in turn hurts the marine economy. In addition, when the total
amount of marine litter exceeds the tolerance limit of marine biological communities, it
will directly affect the virtuous cycle of the marine ecosystem, leading to the deterioration
of the ecological environment, putting biological resources under threat and eventually
causing damage to the marine life ecological environment. In order to address the root
cause of the problem, it is necessary to prevent the occurrence of litter and microplastic
discharges into the ocean in the long term. Therefore, a correct understanding of the types
of marine litter and litter cleanup can help reduce the impact of marine litter on the marine
ecological environment. This carries an important practical significance for future marine
ecological environment management and sustainable development [
1
]. Based on protecting
the marine ecosystem, various methods of marine litter cleanup have been proposed and
Sensors 2021, 21, 6391. https://doi.org/10.3390/s21196391 https://www.mdpi.com/journal/sensors
