Agricultural science and technology: the fishing underwater camera controls the feeding time, and the AI system makes the shrimp fat 30%.

The research team of National Sun Yat-sen University published the "AI Intelligent Culture Achievement Presentation Conference" on the 23rd. Using underwater cameras to record shrimp activity and automatic bait spraying facilities, shrimp can gain 3 grams per week, which is 30% fatter than traditional shrimp culture methods. At present, 2 shrimp farms have installed underwater cameras and bait spraying facilities. In the next three years, the research team will challenge outdoor farms to improve the recognition ability of underwater cameras and become the sharpest eyes of farmers underwater.

Hong Qingzhang, a professor at the Department of Marine Science at National Sun Yat-sen University, said that after a year of experiments, the research team developed a central sewage system that can be activated according to dissolved oxygen parameters, as well as an underwater photography system and bait spraying equipment combined with AI identification. Farmers have the highest degree of inquiry about underwater photography and bait spraying systems.

The underwater camera system monitors shrimp movements all day long by visible and infrared technology, and can be combined with AI identification and automatic bait spraying equipment. When the image identifies that the bait density in the feeding area is lower than 15% or other set values, the bait spraying equipment can be remotely controlled or activated by the system. On the one hand, it can reduce the water pollution caused by bait, and on the other hand, it can accurately grasp the feeding time of bait. Hung says that after six months of testing, shrimp raised in the school's seawater culture pond through this system grew from two grams to three grams per week, 30 percent fatter than shrimp raised in traditional shrimp ponds.

Hong Qingzhang, who jumped from water quality research to shrimp farming industry, said that water quality control is the key to shrimp farming industry. When the water is good, bacteria and viruses are not easy to find shrimps. The research team uses underwater cameras to assist AI identification of accurate bait feeding, and integrates central sewage facilities of variable frequency air compressors to increase the growth rate of grass shrimp to 70%, and can also save 30% of electricity compared with traditional water trucks.

However, the underwater camera system has not yet been tested in all outdoor culture ponds. Hung explains that because there are different algae ecosystems in outdoor culture ponds, the underwater camera's ability to record images and the AI's ability to identify them are all tests. Professor Huang Yingzhe of Zhongshan Department of Finance and Engineering said that the team is also developing artificial intelligence chip modules to improve the performance of the identification system, reduce the power consumption of the underwater imaging system, and meet the user's network bandwidth and data storage requirements.