1,329
Views
0
CrossRef citations to date
0
Altmetric
Research Article

Counting and measuring the size and stomach fullness levels for an intelligent shrimp farming system

, , , , , , & show all
Article: 2268878 | Received 30 May 2023, Accepted 05 Oct 2023, Published online: 18 Oct 2023

Figures & data

Figure 1. Experimental environment. (a) Sketch of the pond. (b) The actual pond.

Figure 1. Experimental environment. (a) Sketch of the pond. (b) The actual pond.

Figure 2. Overall process flow of the proposed method.

Figure 2. Overall process flow of the proposed method.

Figure 3. Network architecture of AR-YOLOv5.

Figure 3. Network architecture of AR-YOLOv5.

Figure 4. Training process of AR-YOLOv5.

Figure 4. Training process of AR-YOLOv5.

Figure 5. The side length definition.

Figure 5. The side length definition.

Figure 6. Comparison of the size estimation methods between regular YOLOv5 and the proposed method in the overlapping shrimps. (a) Regular YOLOv5. (b)(c) Regular YOLOv5 failed to estimate size by segmentation. (d) Our angle detection-based size estimation result.

Figure 6. Comparison of the size estimation methods between regular YOLOv5 and the proposed method in the overlapping shrimps. (a) Regular YOLOv5. (b)(c) Regular YOLOv5 failed to estimate size by segmentation. (d) Our angle detection-based size estimation result.

Table 1. Degree of stomach fullness mapping to Full/Empty.

Figure 7. Inner bounding box and Outer bounding box of shrimp were used to calculate the stomach fullness levels.

Figure 7. Inner bounding box and Outer bounding box of shrimp were used to calculate the stomach fullness levels.

Figure 8. Highlight of the shrimp’s digestive tract segmented using Otsu’s thresholding method.

Figure 8. Highlight of the shrimp’s digestive tract segmented using Otsu’s thresholding method.

Figure 9. Loss function value curve: Comparison of the Loss for five different AR-YOLOv5 models on the training set.

Figure 9. Loss function value curve: Comparison of the Loss for five different AR-YOLOv5 models on the training set.

Figure 10. PR curve: Performance comparison of the five different AR-YOLOv5 models on the testing set.

Figure 10. PR curve: Performance comparison of the five different AR-YOLOv5 models on the testing set.

Table 2. Comparison of different models with data preprocessing.

Table 3. Comparison of different models without data preprocessing.

Figure 11. A four-months growth Shrimps length estimation and its error percentage.

Figure 11. A four-months growth Shrimps length estimation and its error percentage.

Table 4. Absolute measurement error and error percentage of shrimp’s quantity.

Figure 12. Example of stomach fullness levels. Top row: Stomach fullness level < 12.5%. Middle row: 12.5% Stomach fullness level < 50.0%. Bottom row: Stomach fullness level 50.0%.

Figure 12. Example of stomach fullness levels. Top row: Stomach fullness level < 12.5%. Middle row: 12.5% ≤ Stomach fullness level < 50.0%. Bottom row: Stomach fullness level ≥50.0%.

Table 5. Evaluation metric results of two different methods.