Skip to content

Underwater Open Source AI Camera Manual

I. Product Overview

This device is an underwater open source AI camera specifically designed for complex underwater environments. It features a built-in high-performance edge computing module and supports real-time image acquisition, intelligent algorithm processing, and secondary development. This guide aims to assist users in rapid deployment and optimization of monitoring parameters for specific environments.

Rear Interfaces:

  • Port B:​ Connects to the host computer via an 8-core seawater-resistant cable with bulkhead connector. Standard length: 2 m. Optional lengths up to 70 m available; for longer distances, please contact our support team.
  • Port A (Spare):​ Reserved for auxiliary lighting connection.

II. Wiring

Camera Wiring (Without Light)

Wire ColorDefinition
BlackGND
RedPower Input 5.5V ~ 26VDC
BlueN/A
Blue/WhiteN/A
GreenEthernet Tx-
Green/WhiteEthernet Tx+
OrangeEthernet Rx-
Orange/WhiteEthernet Rx+

Camera Wiring (With Light)

Wire ColorDefinition
BlackGND
RedPower Input 12VDC
BlueGND
Blue/WhitePower Input 5.5V ~ 26VDC
GreenEthernet Tx-
Green/WhiteEthernet Tx+
OrangeEthernet Rx-
Orange/WhiteEthernet Rx+

III. Quick Start

Applicable to first-time users or those who only need to view real-time monitoring footage:

1. Hardware Connection: Connect the camera power supply and link the device directly to the computer's network port via an Ethernet cable.

2. Network Configuration:

  • Set the computer's Ethernet IPv4 address to: 192.168.2.1 (Subnet mask: 255.255.255.0).

3. Video Viewing: Open a browser and enter the following addresses to access:

Ⅳ. Secondary Development and Parameter Optimization Instructions

For underwater environments (e.g., suspended particles, ripple interference), developers can log in via SSH to modify algorithm parameters and improve monitoring accuracy.

1. Remote Login: Open the Windows PowerShell command line window on your computer and enter the following commands:

2. Parameter Configuration File Modification: After logging in, open the configuration file using the nano editor:

  • Operation Command: sudo nano config.ini

3. Camera Parameter Configuration and Tuning Guide: To achieve optimal monitoring results, please modify the parameters in the config.ini file according to actual scenario requirements.

ModuleParameter NameAdjustment LogicSuggested Direction
Feature Switchenable_...Set to true to enable, false to disable.Enable/disable as needed
Brightness Adjustmentgamma_valIncrease if the image is too dark; decrease if it is too bright.Based on ambient light
Edge Detectionblur_kernelMust be an odd number. Increase if the image is too noisy/grainy; decrease if it is too blurry.3, 5, 7, 9, 11...
Edge Detectioncanny_lowIncrease if there are too many noise points; adjust canny_high accordingly.Increase to filter noise
Motion Detectionmotion_thresholdSensitivity threshold. Increase to reduce false alarms; decrease to reduce missed detections.50 (Sensitive) - 120 (Robust)
Motion Detectionmin_motion_areaArea filtering. Increase to filter out small targets; decrease to capture small targets.Based on target size

⚠️ Developer Tip: After modifying the configuration file, save the file (Ctrl + O), exit the editor (Ctrl + X), and refresh the browser to apply the configuration. For complex underwater environments, it is recommended to prioritize adjusting blur_kernel and min_motion_area to suppress false alarms caused by environmental fluctuations.