Abstract
Object counting in aquaculture is an important task, and has been widely applied on fish population estimation, lobster abundance estimation, and scallop stock, etc. However, underwater object counting is challenging for biologists and marine scientists, because of the diversity of background of the lake or ocean, the uncertainty of the object motion, and the occlusion between objects. With the rapid development of sensor, computer vision and acoustic technologies, advanced and efficient counting methods are available in aquaculture. We reviewed underwater object counting methods in aquaculture, provided a survey including more than 50 papers in the recent 10 years, and analyzed the pros and cons of the counting methods and the applicable scenarios of those methods. Finally, the major challenges and future trends of underwater object counting in aquaculture are discussed.