An improved algae-YOLO model based on deep learning for object detection of ocean microalgae considering aquacultural lightweight deployment

Abstract

Algae are widely distributed and have a considerable impact on water quality. Harmful algae can degrade water quality and be detrimental to aquaculture, while beneficial algae are widely used. The accuracy and speed of existing intelligent algae detection methods are available, but the size of parameters of models is large, the equipment requirements are high, the deployment costs are high, and there is still little research on lightweight detection methods in the area of algae detection. In this paper, we propose an improved Algae-YOLO object detection approach, which is based on ShuffleNetV2 as the YOLO backbone network to reduce the parameter space, adapting the ECA attention mechanism to improve detection accuracy, and redesigning the neck structure replacing the neck structure with ghost convolution module for reducing the size of parameters, finally the method achieved the comparable accuracy. Experiments showed that the Algal-YOLO approach in this paper reduces the size of parameters by 82.3%, and the computation (FLOPs) is decreased from 16G to 2.9G with less loss of accuracy, and mAP by only 0.007 when compared to the original YOLOv5s. With high accuracy, the smaller model size are achieved, which reduces the equipment cost during actual deployment and helps to promote the practical application of algae detection.

ADNet: A Real-Time Floating Algae Segmentation Using Distillation Network

“…Park et al [22] proposed to utilize the you only look once (YOLO) [23] model for algal image detection, achieving a balance between inference time and accuracy. Liu et al [24] introduced an enhanced version of the algae-YOLO approach, using the ShuffleNetV2 [25] as the backbone network to reduce the parameter space.…”

SenseLite: A YOLO-Based Lightweight Model for Small Object Detection in Aerial Imagery

“…Moreover, it integrates an α-CIoU loss method to mitigate the bias between favorable and unfavorable samples in skyborne images [12]. Zhao, Lei, and their team proposed an innovative and lightweight method for spotting objects in aerial images, LAI-YOLOv5s, which demonstrated both high accuracy and reduced computational demands [13].…”

Detection of algal tiny objects based on morphological features

“…The image classifier performs well in the field of algae identification 3,13‐15 . At present, a variety of object detection frameworks have been applied to algae identification, 16‐18 which replaces the traditional microscopy method and improves the efficiency of classification and identification 19‐22 . For example, Cao et al 19 used the improved YOLOv3 model of MobileNet 23 and SPP 24 to achieve better recognition performance on two single‐category microalgae datasets, indicating that the object detection model can use the image features of algae not rich for recognition.…”

“…traditional microscopy method and improves the efficiency of classification and identification. [19][20][21][22] For example, Cao et al 19 used the improved YOLOv3 model of MobileNet 23 and SPP 24 to achieve better recognition performance on two single-category microalgae datasets, indicating that the object detection model can use the image features of algae not rich for recognition. Park et al 21 applied the YOLOv3 model with Darknet53 25 as the backbone to train on a data set with a limited number of algae, which proved that the lightweight object detection model has more advantages in such tasks.…”