Integrate MSRCR and Mask R-CNN to Recognize Underwater Creatures on Small Sample Datasets

Abstract

The poor quality of optical imaging caused by the complex and varying underwater environment is a significant challenge to underwater target recognition. Moreover, the insufficiency of relevant datasets may lead to the overfitting problem in target recognition models based on deep learning. Taking the instance segmentation of three underwater creatures (echinus, holothurian, and starfish) as an example, we propose a new method for recognition of underwater creatures. It combines the MSRCR (multi-scale Retinex with color restoration) image enhancement algorithm and the Mask R-CNN (regionbased convolutional neural work) framework, and achieves a mAP (mean average accuracy) value higher than 90% on a small sample dataset. This method consists of three major steps. First, the dataset with 84 images is augmented (flip, adding noise, and GAN (generative adversarial networks)) to 430 images, and all images are enhanced with MSRCR to improve their qualities; Second, the model is pre-trained on the COCO (Microsoft common objects in context) dataset to shorten the training time and overcome overfitting; Finally, the pre-trained model is transferred to the underwater dataset, and the whole training process is completed. We achieve 97.46% precision and 94.52% recall, and the mAP (intersection over union (IOU) =50) is 94.84%. The effectiveness of the proposed method is verified by comparing it with several popular target recognition models, including SSD (Single Shot Detector), YOLOv3 (You only look once), original Mask R-CNN, and a SIFT-based (Scale-invariant feature transform) model.

Distributed Deep Learning and Energy-Efficient Real-Time Image Processing at the Edge for Fish Segmentation in Underwater Videos

“…Besides, seawater reacts differently to different light spectra, according to its frequency-dependent power absorption. These and many other extreme imaging limitations make the underwater environment a nonuniform imaging medium [12]. These nonuniform image degradation processes make underwater image segmentation a challenging task.…”

“…One of the most recent works in underwater image segmentation combines the multi-scale Retinex image enhancement algorithm and the Mask R-CNN framework to recognize marine echinoderm [12]. While it is limited to echinus, holothurian, and starfish, the work in [12] can be expanded to incorporate more sea creatures. Besides, the trained model is big and consumes high energy for inferencing.…”

Research on Marine Life Detection Algorithm Based on Deep Learning

“…After continuous exploration, the common underwater image enhancement algorithms now include the method based on dark channel prior, the method based on histogram equalization and the method based on Retinex. Figure 1 (10) shows the original image and the resulting images after processing by these three methods. Table 1 (10) shows the impact of these three image enhancement algorithms on the target detection and recognition results.…”

“…Figure 1 (10) shows the original image and the resulting images after processing by these three methods. Table 1 (10) shows the impact of these three image enhancement algorithms on the target detection and recognition results. Fig.…”

A New Real-Time Detection and Tracking Method in Videos for Small Target Traffic Signs

“…Moreover, the primary detection object is small targets, which lead to even more undesirable results. According to the Vicinal Risk Minimization (VRM) principle, the generalization capacity should be improved by creating samples similar to the training samples for data augmentation [47]. Therefore, data augmentation is employed for small sample traffic signs in the TT100K to achieve the purpose of sample equilibrium, which is inspired by [48].…”