Semantic segmentation method of underwater images based on encoder-decoder architecture

Abstract

With the exploration and development of marine resources, deep learning is more and more widely used in underwater image processing. However, the quality of the original underwater images is so low that traditional semantic segmentation methods obtain poor segmentation results, such as blurred target edges, insufficient segmentation accuracy, and poor regional boundary segmentation effects. To solve these problems, this paper proposes a semantic segmentation method for underwater images. Firstly, the image enhancement based on multi-spatial transformation is performed to improve the quality of the original images, which is not common in other advanced semantic segmentation methods. Then, the densely connected hybrid atrous convolution effectively expands the receptive field and slows down the speed of resolution reduction. Next, the cascaded atrous convolutional spatial pyramid pooling module integrates boundary features of different scales to enrich target details. Finally, the context information aggregation decoder fuses the features of the shallow network and the deep network to extract rich contextual information, which greatly reduces information loss. The proposed method was evaluated on RUIE, HabCam UID, and UIEBD. Compared with the state-of-the-art semantic segmentation algorithms, the proposed method has advantages in segmentation integrity, location accuracy, boundary clarity, and detail in subjective perception. On the objective data, the proposed method achieves the highest MIOU of 68.3 and OA of 79.4, and it has a low resource consumption. Besides, the ablation experiment also verifies the effectiveness of our method.

Deep Learning for Enhanced Marine Vision: Object Detection in Underwater Environments

“…Villon et al [18] utilized a deep learning model for the detection of coral reef fishes. Real-time detection demands were also fulfilled by Yang et al [19], employing the YOLOv3 framework [20] for underwater object detection. Additionally, [18]employed the Fast-RCNN framework for fish species detection, later adopting Faster-RCNN [21] to optimize fish detection speed.…”

“…Semantic segmentation for underwater object detection is a challenging computer vision task that involves the accurate classification and delineation of various objects and regions within underwater imagery [16]. It plays a critical role in understanding the complex underwater environment and has significant applications in marine research, environmental monitoring, underwater robotics, and ocean exploration [17,19].…”

Improving Semantic Segmentation Performance in Underwater Images

“…In addition, traditional methods are generally not very transferable or robust, so the segmentation result of a single traditional method is poor in most cases [1]. It is therefore necessary to resort to advanced approaches, often involving Deep Learning [2], to better address these underwater challenges [3][4][5]. For these methods to work well for this type of task, which requires training models, appropriate datasets must be used.…”

Underwater Fish Segmentation Algorithm Based on Improved PSPNet Network

“…In 2021, Wenbo Zhang et al [9] proposed a two-pool polymerization attentional network to improve underwater fish segmentation accuracy using a pool polymerization positional attention module and a pool polymerization channel attention module. In 2022, Jinkang Wang et al [10] proposed an underwater image semantic segmentation method to precisely segment targets; however, the first step in this method was to improve image quality by performing image enhancement operations based on multispatial transformation. In recent years, increasing numbers of researchers have begun to improve segmentation accuracy from the perspective of integrating multiscale features of fish targets, such as the multiscale CNN network [11][12][13][14] and the porous GAN network [15][16][17][18].…”