SharkSpotter: Shark Detection with Drones for Human Safety and Environmental Protection

Sharma, Nabin; Saqib, Muhammed; Scully‐Power, Paul; Blumenstein, Michael

doi:10.1007/978-3-030-72188-6_11

Cited by 9 publications

(10 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The technological capabilities of drones are rapidly improving, offering new opportunities to build on the research presented here and optimise the use of drones for detecting sharks in the coastal environment. Artificial Intelligence (AI) is one area which is currently being explored to assess whether it can be used to improve the detection rates of sharks and improve operational efficiency [ 32 , 46 , 47 , 48 ]. Recent testing of the SharkSpotter © deep-learning based AI system in NSW, showed that it can achieve a detection success of 90%, which is typically higher than the pilot because they can be distracted by other objects and they are concentrating on flying the drone [ 47 , 48 ].…”

Section: Discussionmentioning

confidence: 99%

“…Artificial Intelligence (AI) is one area which is currently being explored to assess whether it can be used to improve the detection rates of sharks and improve operational efficiency [ 32 , 46 , 47 , 48 ]. Recent testing of the SharkSpotter © deep-learning based AI system in NSW, showed that it can achieve a detection success of 90%, which is typically higher than the pilot because they can be distracted by other objects and they are concentrating on flying the drone [ 47 , 48 ]. This AI system operates in near real time and flags any objects detected, which helps to reduce reliance on pilots to observe sharks, improving efficiency and reducing pilot fatigue [ 48 ].…”

Section: Discussionmentioning

confidence: 99%

“…Recent testing of the SharkSpotter © deep-learning based AI system in NSW, showed that it can achieve a detection success of 90%, which is typically higher than the pilot because they can be distracted by other objects and they are concentrating on flying the drone [ 47 , 48 ]. This AI system operates in near real time and flags any objects detected, which helps to reduce reliance on pilots to observe sharks, improving efficiency and reducing pilot fatigue [ 48 ]. Furthermore, Purcell et al, [ 29 ] had similar detection rates but also isolated animals into species categories, further enhancing the capabilities and utility for use in beach safety programs.…”

Section: Discussionmentioning

confidence: 99%

See 2 more Smart Citations

Factors Affecting Shark Detection from Drone Patrols in Southeast Queensland, Eastern Australia

Mitchell

Scott-Holland

Butcher

2022

Biology

View full text Add to dashboard Cite

Drones enable the monitoring for sharks in real-time, enhancing the safety of ocean users with minimal impact on marine life. Yet, the effectiveness of drones for detecting sharks (especially potentially dangerous sharks; i.e., white shark, tiger shark, bull shark) has not yet been tested at Queensland beaches. To determine effectiveness, it is necessary to understand how environmental and operational factors affect the ability of drones to detect sharks. To assess this, we utilised data from the Queensland SharkSmart drone trial, which operated at five southeast Queensland beaches for 12 months in 2020–2021. The trial conducted 3369 flights, covering 1348 km and sighting 174 sharks (48 of which were >2 m in length). Of these, eight bull sharks and one white shark were detected, leading to four beach evacuations. The shark sighting rate was 3% when averaged across all beaches, with North Stradbroke Island (NSI) having the highest sighting rate (17.9%) and Coolum North the lowest (0%). Drone pilots were able to differentiate between key shark species, including white, bull and whaler sharks, and estimate total length of the sharks. Statistical analysis indicated that location, the sighting of other fauna, season and flight number (proxy for time of day) influenced the probability of sighting sharks.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Factors Affecting Shark Detection from Drone Patrols in Southeast Queensland, Eastern Australia

Mitchell

Scott-Holland

Butcher

2022

Biology

View full text Add to dashboard Cite

show abstract

“…The Little Ripper Group 2 deployed one of the first such drone-based shark spotting systems in Australia, utilising desktop-class computing hardware to detect sharks in streaming video. Their object detection model was developed by researchers at the University of Technology Sydney who reported a mean average precision of mAP >90 % when tested on an internal dataset (Sharma et al, 2018;Sharma et al, 2022). However, the early system proved unreliable when deployed to new locations and the computing hardware was not portable or robust enough for the harsh beach environment (private communication, Surf Life Saving New South Wales).…”

Section: Machine Learning On Beachesmentioning

confidence: 99%

Assessing the ability of deep learning techniques to perform real-time identification of shark species in live streaming video from drones

Purcell

Walsh²,

Colefax³

et al. 2022

Front. Mar. Sci.

View full text Add to dashboard Cite

Over the last five years remotely piloted drones have become the tool of choice to spot potentially dangerous sharks in New South Wales, Australia. They have proven to be a more effective, accessible and cheaper solution compared to crewed aircraft. However, the ability to reliably detect and identify marine fauna is closely tied to pilot skill, experience and level of fatigue. Modern computer vision technology offers the possibility of improving detection reliability and even automating the surveillance process in the future. In this work we investigate the ability of commodity deep learning algorithms to detect marine objects in video footage from drones, with a focus on distinguishing between shark species. This study was enabled by the large archive of video footage gathered during the NSW Department of Primary Industries Drone Trials since 2016. We used this data to train two neural networks, based on the ResNet-50 and MobileNet V1 architectures, to detect and identify ten classes of marine object in 1080p resolution video footage. Both networks are capable of reliably detecting dangerous sharks: 80% accuracy for RetinaNet-50 and 78% for MobileNet V1 when tested on a challenging external dataset, which compares well to human observers. The object detection models correctly detect and localise most objects, produce few false-positive detections and can successfully distinguish between species of marine fauna in good conditions. We find that shallower network architectures, like MobileNet V1, tend to perform slightly worse on smaller objects, so care is needed when selecting a network to match deployment needs. We show that inherent biases in the training set have the largest effect on reliability. Some of these biases can be mitigated by pre-processing the data prior to training, however, this requires a large store of high resolution images that supports augmentation. A key finding is that models need to be carefully tuned for new locations and water conditions. Finally, we built an Android mobile application to run inference on real-time streaming video and demonstrated a working prototype during fields trials run in partnership with Surf Life Saving NSW.

show abstract

“…Consequently, artificial intelligence techniques have been utilized in the provisioning of civilian and military services [2]. In this context, drones have been adopted for express shipping and delivery [3][4][5], natural disaster prevention [6,7], geographical mapping [8], search and rescue operations [9], aerial photography for journalism and film making [10], providing essential materials [11], border control surveillance [12], and building safety inspection [13]. Even though drone technology offers a multitude of benefits, it raises mixed concerns when it comes to how it will be used in the future.…”

Section: Introductionmentioning

confidence: 99%

Simultaneous Tracking and Recognizing Drone Targets with Millimeter-Wave Radar and Convolutional Neural Network

Solaiman,

Alsuwat,

Alharthi

2023

ASI

View full text Add to dashboard Cite

In this paper, a framework for simultaneous tracking and recognizing drone targets using a low-cost and small-sized millimeter-wave radar is presented. The radar collects the reflected signals of multiple targets in the field of view, including drone and non-drone targets. The analysis of the received signals allows multiple targets to be distinguished because of their different reflection patterns. The proposed framework consists of four processes: signal processing, cloud point clustering, target tracking, and target recognition. Signal processing translates the raw collected signals into spare cloud points. These points are merged into several clusters, each representing a single target in three-dimensional space. Target tracking estimates the new location of each detected target. A novel convolutional neural network model was designed to extract and recognize the features of drone and non-drone targets. For the performance evaluation, a dataset collected with an IWR6843ISK mmWave sensor by Texas Instruments was used for training and testing the convolutional neural network. The proposed recognition model achieved accuracies of 98.4% and 98.1% for one and two targets, respectively.

show abstract

SharkSpotter: Shark Detection with Drones for Human Safety and Environmental Protection

Cited by 9 publications

References 23 publications

Factors Affecting Shark Detection from Drone Patrols in Southeast Queensland, Eastern Australia

Factors Affecting Shark Detection from Drone Patrols in Southeast Queensland, Eastern Australia

Assessing the ability of deep learning techniques to perform real-time identification of shark species in live streaming video from drones

Simultaneous Tracking and Recognizing Drone Targets with Millimeter-Wave Radar and Convolutional Neural Network

Contact Info

Product

Resources

About