AI-Powered Fish Survey for Biodiversity & Biomass Estimation

DCube developed an AI-powered underwater video analytics system to automatically detect, classify, and count fish species from stereo camera footage. The solution enables scalable biodiversity assessment and biomass estimation to support marine conservation and sustainable resource management.

The Client

Client Name: Mapizy
Industry: Conservation
Region: Australia
Company Size: Enterprise

The Challenge

Marine ecology research generates vast volumes of underwater video data that are impractical to analyze manually. Fish species vary significantly in appearance, size, and movement, while underwater conditions such as low visibility, lighting variation, and occlusions make accurate detection and classification extremely challenging—especially at scale.

The Solution

DCube designed a comprehensive computer vision pipeline combining fish detection, taxonomy classification, tracking, and stereo-based measurement techniques. The system processes underwater videos to automatically identify species, maintain accurate counts, and estimate fish biomass, enabling efficient and non-invasive marine surveys.

Key Features

  • Automated fish detection and species classification
  • Taxonomy-aware classification models for biodiversity analysis
  • Multi-object tracking to avoid double counting
  • Biomass estimation using stereo vision and size measurements
  • Robust performance under challenging underwater conditions

Technologies Used

  • Computer Vision (Detection, Tracking, Stereo Vision)
  • Deep Learning Models for Species Classification
  • Python
  • Custom AI Processing Pipelines

Results & Impact

  • Enabled large-scale, automated biodiversity surveys
  • Reduced manual annotation and analysis effort for researchers
  • Supported preservation of endangered fish species through accurate population insights
  • Provided a non-invasive, low-cost alternative to traditional marine survey methods
  • Improved data quality for sustainable fisheries and ecological research

Ready to Transform Your Business?

Let's discuss how we can help you achieve similar results.