Shrimp sound detection system and detection method

Abstract

A shrimp sound detection system is disclosed for acquiring the feeding sounds of shrimp in a water area and controlling the feeding action of a feeding device based on these sounds. The system includes a sound sensing module, a feeding device, and a detection device. The sound sensing module is positioned within the water area and acquires the feeding sounds of the shrimp, converting them into sound signals. The feeding device is positioned at the outer edge of the water area and is used to feed the shrimp. The detection device uses a feature recognition model to extract the feeding sound characteristics of the shrimp from the sound signals, generates sound feature signals based on these characteristics, then generates sound intensity grading based on the sound feature signals, and finally generates control signals based on the sound intensity grading. These control signals are then transmitted to the feeding device to control its feeding action.

Claims

1. A shrimp sound detection system for acquiring a feeding sound of a group of shrimp (S) in a body of water (W) and controlling a feeding action of a feeding device (2) based on the feeding sound, the shrimp sound detection system (100) comprising: a sound sensing module (1) disposed in the body of water (W) and acquiring the feeding sound of the group of shrimp (S) in the body of water (W) at specific intervals, and converting the feeding sound into a sound signal; a feeding device (2) disposed at an outer edge of the body of water (W) and used to feed a plurality of feeds into the body of water (W); and a detection device (3) signal-connected to the sound sensing module (1) and the feeding device (2), comprising: An audio recognition module (33) uses a feature recognition model (331) to extract multiple feeding sound features of the shrimp group (S) when feeding from the sound signal, and then generates a sound feature signal based on the multiple feeding sound features; and a control module (35) is connected to the audio recognition module (33) to generate a control signal based on the sound feature signal, and transmits the control signal to the feeding device (2) to control the feeding action of the feeding device (2).

2. The shrimp sound detection system as described in claim 1, wherein the detection device (3) further includes a feeding intensity analysis module (34), which is signal-connected to the audio recognition module (33) and the control module (35) to generate a sound intensity rating based on the plurality of feeding sound features transmitted by the audio recognition module (33), and then transmits the sound intensity rating to the control module (35).

3. The shrimp sound detection system as described in claim 1 or claim 2, wherein the detection device (3) further includes a filter module (31) connected to the sound sensing module (1), the filter module (31) being used to perform a filtering process on the sound signal acquired by the sound sensing module (1) to remove noise from the sound signal.

4. The shrimp sound detection system as described in claim 3, wherein the detection device (3) further includes a spectrum conversion module (32) connected to the filtering module (31), the spectrum conversion module (32) being used to perform a spectrum conversion on the sound signal that has been filtered, so as to make the plurality of feeding sound features of the shrimp group (S) in the sound signal more obvious, thereby improving the accuracy of the audio recognition module (33) in extracting the plurality of feeding sound features.

5. The shrimp sound detection system as described in claim 4, wherein the spectrum conversion process is a Mel spectrum conversion or a Fourier transformation.

6. The shrimp sound detection system as described in claim 2, wherein the detection device (3) further includes an integrated report module (36) connected to the feeding intensity analysis module (34), the integrated report module (36) being used to compile all relevant signal data into a report.

7. The shrimp sound detection system as described in claim 6, wherein the detection device (3) further includes a display module (37) signal-connected to the integrated report module (36), the display module (37) being used to display the report output by the integrated report module (36) to a user for reference.

8. The shrimp sound detection system as described in claim 1, wherein a feeding tray (F) is provided in the water area (W), the feeding tray (F) is used to collect the plurality of feeds fed by the feeding device (2) to concentrate the plurality of feeds, and the sound sensing module (1) is located directly above the feeding tray (F).

9. A shrimp sound detection method, applied in a body of water (W), for acquiring and detecting a feeding sound of a group of shrimp (S), the method comprising the following steps: acquiring the feeding sound of the group of shrimp (S) in the body of water (W) through a sound sensing module (1) installed in the body of water (W), and converting the feeding sound into a sound signal; extracting a plurality of feeding sound features of the group of shrimp (S) feeding from the sound signal through an audio recognition module (33) using a feature recognition model (331), and generating a sound feature signal based on the plurality of feeding sound features; generating a control signal based on the sound feature signal through a control module (35); and feeding or stopping feeding in the body of water (W) according to the control signal through a feeding device (2).

10. The shrimp sound detection method as described in claim 9 further includes the following steps: generating a sound intensity classification based on the sound characteristic signal through a feeding intensity analysis module (34), and then transmitting it to the control module (35).

11. The shrimp sound detection method as described in claim 9 or claim 10, further comprising: performing a filtering process on the sound signal acquired by the sound sensing module (1) through a filtering module (31) to remove noise from the sound signal.

12. The shrimp sound detection method as described in claim 11 further includes the following steps: performing a spectrum conversion process on the sound signal that has been filtered by a spectrum conversion module (32) to make the feeding sound features of the shrimp group (S) in the sound signal more obvious, so as to improve the accuracy of the audio recognition module (33) in extracting the feeding sound features.

13. The shrimp sound detection method as described in claim 12, wherein the spectral conversion process is a Mel Spectrogram or a Fourier Transform.

14. The shrimp sound detection method as described in claim 10 further includes the following steps: compiling all relevant signal data into a report through an integrated report module (36).

15. The shrimp sound detection method as described in claim 14 further includes the following steps: displaying the report output by the integrated report module (36) to a user for reference through a display module (37).

Images