Centre for Optical and Electromagnetic Research, South China Academy of Advanced Optoelectronics, South China Normal University, Guangzhou 510006, China.
Sensors (Basel). 2024 Sep 30;24(19):6328. doi: 10.3390/s24196328.
Counting shrimp larvae is an essential part of shrimp farming. Due to their tiny size and high density, this task is exceedingly difficult. Thus, we introduce an algorithm for counting densely packed shrimp larvae utilizing an enhanced You Only Look Once version 5 (YOLOv5) model through a regional segmentation approach. First, the C2f and convolutional block attention modules are used to improve the capabilities of YOLOv5 in recognizing the small shrimp. Moreover, employing a regional segmentation technique can decrease the receptive field area, thereby enhancing the shrimp counter's detection performance. Finally, a strategy for stitching and deduplication is implemented to tackle the problem of double counting across various segments. The findings from the experiments indicate that the suggested algorithm surpasses several other shrimp counting techniques in terms of accuracy. Notably, for high-density shrimp larvae in large quantities, this algorithm attained an accuracy exceeding 98%.
对虾苗进行计数是对虾养殖的重要环节。由于虾苗个体微小且密度较高，这项工作极具挑战性。因此，我们提出了一种利用增强型 You Only Look Once 版本 5（YOLOv5）模型，通过区域分割方法对密集的虾苗进行计数的算法。首先，使用 C2f 和卷积块注意力模块提高 YOLOv5 识别小对虾的能力。此外，采用区域分割技术可以减少感受野面积，从而提高虾计数器的检测性能。最后，采用拼接和去重策略来解决不同区域重复计数的问题。实验结果表明，所提出的算法在准确性方面优于其他几种虾苗计数技术。特别是在大量高密度虾苗的情况下，该算法的准确率超过 98%。
