Development of an underwater camera system for inland freshwater aquaculture
Computer vision and image processing technologies are applied towards aquatic research to understand fish and its interaction with other fishes and their environment. The understanding of vision-based data acquisition and processing aids in developing predictive frameworks and decision support syste...
Saved in:
| Summary: | Computer vision and image processing technologies are applied towards aquatic research to understand fish and its interaction with other fishes and their environment. The understanding of vision-based data acquisition and processing aids in developing predictive frameworks and decision support systems for efficient aquaculture monitoring and management. However, this emerging field is confronted by a lack of high-quality underwater visual data, whether from public or local setups. An accessible underwater camera system that intensively obtains underwater visual data periodically and in real-time is the most desired system for such emerging studies. In this regard, an underwater camera system that captures underwater images from an inland freshwater aquaculture setup was proposed. The components of the underwater camera system are primarily based on Raspberry Pi, an open-source computing platform. The underwater camera continuously provides a real-time video streaming link of underwater scenes, and the local processor periodically acquires and stores data from this link in the form of images. These data are stored locally and remotely. Also, the local processor initiates a connection to a remote processor to allow the remote view of the real-time video streaming link. Aside from accessing the data and streaming link remotely, the remote processor analyzes the statistics of the underwater images to motivate the application of color balance and fusion, a state-of-the-art underwater image enhancement method. The applications of the proposed system and the enhancement to the captures are objectively evaluated. The proposed system captured around 1.2 Gb worth of 8 MP underwater images during daytime every day and stored these images in cloud storage. Also, the system captured subjects within 10-35 cm of turbid fishpond water. The statistical analysis of the gathered data revealed that underwater images from turbid fishpond setups have low quality in terms of inaccurate color representations (i.e., dominant green intensities and mostly submissive blue intensities) and low contrast. These observations appropriated the application of color balance and fusion to the locally acquired data. Furthermore, the objective evaluation revealed that color balance and fusion is the most effective method of improving information content and edge details, as quantified by high color information entropies and high average gradients. These metrics revealed the effectiveness of the proposed data acquisition and preprocessing system. |
|---|