Anomalous Behaviors Detection for Underwater Fish Using AI Techniques

Abstract

Anomalous events detection in real-world video scenes is a challenging problem owing to the complexity of anomaly and the untidy backgrounds and objects in the scenes. Although there are already many studies on dealing with this problem using deep neural networks, very little literature aims for real-time detection of the anomalous behavior of fish. This paper presents an underwater fish anomalous behavior detection method by combining deep learning object detection, DCG (Directed Cycle Graph), fish tracking, and DTW (Dynamic Time Warping). The method is useful for detecting the biological anomalous behavior of underwater fish in advance so that early countermeasures can be planned and executed. Also, through post-analysis it is possible to access the cause of diseases or death, so as to reduce unnecessary loss, facilitate precision breeding selection, and achieve ecological conservation education as well. A smart aquaculture system incorporating the proposed method and IoT sensors allows extensive data collection during the system operation in various farming fields, thus enabling to develop optimal culturing conditions, both are particularly useful for researchers and the aquaculture industry.

Deep learning for visual recognition and detection of aquatic animals: A review

“…176 DL networks, combined with the Directed Cycle Graph (DCG) and Dynamic Time Warping (DTW), can effectively detect the abnormal behaviour of fish to determine the cause of diseases or death. 177 Therefore, the monitoring of abnormal behaviour of aquatic animals is helpful to take corresponding measures in time to avoid greater losses.…”

“…Large‐scale losses of fisheries are often caused by the disease of individual fish, and mass death of fish is often accompanied by abnormal behaviour in the early stage of fish disease 176 . DL networks, combined with the Directed Cycle Graph (DCG) and Dynamic Time Warping (DTW), can effectively detect the abnormal behaviour of fish to determine the cause of diseases or death 177 . Therefore, the monitoring of abnormal behaviour of aquatic animals is helpful to take corresponding measures in time to avoid greater losses.…”

Identification of Genes Related to Cold Tolerance and Novel Genetic Markers for Molecular Breeding in Taiwan Tilapia (Oreochromis spp.) via Transcriptome Analysis

“…The combination of high-throughput transcriptome sequencing and assembly, database comparison and functional annotation, gene differential expression, and molecular marker genotype analysis was quite effective in exploring the theory of key regulatory pathways in stress physiology, the development of molecular marker-assisted breeding platforms and their potential applications [ 78 , 79 ]. In the future, in addition to further research examining the effects of aquatic diet additives on the cold resistance traits of aquatic products, artificial intelligence, and nutrigenomics research strategies such as the AI-assisted adversity abnormal behavior identification system [ 80 ] will be used. The recently developed whole genomic SNP array technology is also crucial for the follow-up study of more Taiwan tilapia resistance traits and genome-wide association studies, to improve the research and development of precise nutrition breeding regulation mechanisms.…”

Marine pasture monitoring system based on edge computing

“…In the case of a single object and low density, it has a good automatic counting and processing ability, while there are still technical bottlenecks. However, the Marine pasture monitoring system based on edge computing can realize the accurate classification of species, density calculation, and monitoring of abnormal behavior and specific disaster invasion through deep machine learning [13] , which is bound to move towards self-learning and intelligent calculation to a certain extent.…”