Real-time fish detection in complex backgrounds using probabilistic background modelling

Ahmad, Sahrim; Maqbool, Salman; Khan, Abdul Hannan; Jalal, Ahsan; Shafait, Faisal

doi:10.1016/j.ecoinf.2019.02.011

Cited by 57 publications

(23 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Inter-habitat differences might mean that a deep learning model trained on one habitat will not be reliable for others. For instance, structural complexity may influence model performance, as background confusion and foreground camouflage might compromise accuracy (Salman et al 2019a).…”

Section: Introductionmentioning

confidence: 99%

Deep learning for automated analysis of fish abundance: the benefits of training across multiple habitats

Ditria

Sievers

Lopez‐Marcano

et al. 2020

Environ Monit Assess

View full text Add to dashboard Cite

Environmental monitoring guides conservation and is particularly important for aquatic habitats which are heavily impacted by human activities. Underwater cameras and uncrewed devices monitor aquatic wildlife, but manual processing of footage is a significant bottleneck to rapid data processing and dissemination of results. Deep learning has emerged as a solution, but its ability to accurately detect animals across habitat types and locations is largely untested for coastal environments. Here, we produce five deep learning models using an object detection framework to detect an ecologically important fish, luderick (Girella tricuspidata). We trained two models on footage from single habitats (seagrass or reef) and three on footage from both habitats. All models were subjected to tests from both habitat types. Models performed well on test data from the same habitat type (object detection measure: mAP50: 91.7 and 86.9% performance for seagrass and reef, respectively) but poorly on test sets from a different habitat type (73.3 and 58.4%, respectively). The model trained on a combination of both habitats produced the highest object detection results for both tests (an average of 92.4 and 87.8%, respectively). The ability of the combination trained models to correctly estimate the ecological abundance metric, MaxN, showed similar patterns. The findings demonstrate that deep learning models extract ecologically useful information from video footage accurately and consistently and can perform across habitat types when trained on footage from the variety of habitat types.

show abstract

Section: Introductionmentioning

confidence: 99%

Deep learning for automated analysis of fish abundance: the benefits of training across multiple habitats

Ditria

Sievers

Lopez‐Marcano

et al. 2020

Environ Monit Assess

View full text Add to dashboard Cite

show abstract

“…Moreover, the spatiocontextual Gaussian mixture model was used to estimate spatial and temporal correlations. A similar strategy was used in recent research wherein a probabilistic background was established for realtime fish detection [21].…”

Section: Object Detection From Underwater Imagesmentioning

confidence: 99%

Optical Prior‐Based Underwater Object Detection with Active Imaging

Shen

Chen

et al. 2021

Complexity

View full text Add to dashboard Cite

Underwater object detection plays an important role in research and practice, as it provides condensed and informative content that represents underwater objects. However, detecting objects from underwater images is challenging because underwater environments significantly degenerate image quality and distort the contrast between the object and background. To address this problem, this paper proposes an optical prior-based underwater object detection approach that takes advantage of optical principles to identify optical collimation over underwater images, providing valuable guidance for extracting object features. Unlike data-driven knowledge, the prior in our method is independent of training samples. The fundamental novelty of our approach lies in the integration of an image prior and the object detection task. This novelty is fundamental to the satisfying performance of our approach in underwater environments, which is demonstrated through comparisons with state-of-the-art object detection methods.

show abstract

“…Visi komputer dan pendekatan pemrosesan gambar untuk deteksi ikan bawah air mendapatkan perhatian yang khusus oleh para ilmuwan kelautan (Salman, Maqbool, Khan, Jalal, & Shafait, 2019) dengan mengestimasi keberadaan ikan dari video dan gambar dapat mendukung ahli biologi kelautan untuk memahami lingkungan bawah laut yang alami, mempromosikan pelestariannya dan mempelajari perilaku interaksi antara hewan laut yang menjadi bagian darinya (Li, Shang, Qin, & Chen, 2016) Untuk melakukan deteksi ikan dibutuhkan sebuah proses pengenalan ikan. Pengenalan ikan merupakan proses mengidentifikasikan ikan berdasarkan gambaran bentuk pola tubuh ikan beserta ciri-cirinya (Santoso, Setiyono, & Isnanto, 2011).…”

Section: Pendahuluanunclassified

Deteksi Ikan Dengan Menggunakan Algoritma Histogram of Oriented Gradients

Anggraeny

Rahmat

Pratama

2020

JIM

View full text Add to dashboard Cite

Indonesia merupakan negara yang kaya akan sumber daya alam baik hayati maupun non-hayati. Salah satu sumber daya alam hayati yang sangat banyak jumlahnya di Indonesia adalah laut, Untuk mempermudah mengidentifikasikan ikan, dapat memanfaatkan sebuah teknologi yang dapat membantu manusia untuk dapat mengenali ikan dengan menggunakan visi komputer dan pendekatan pemrosesan gambar untuk deteksi ikan dan bukan ikan menggunakan algoritma Histogram of Oriented Gradients (HOG) dan AdaBoost-SVM. Hasil penelitian menunjukkan bahwa metode HOG dan AdaBoost-SVM dapat menghasilkan tingkat akurasi rata-rata sebesar 84.8%.

show abstract

Real-time fish detection in complex backgrounds using probabilistic background modelling

Cited by 57 publications

References 15 publications

Deep learning for automated analysis of fish abundance: the benefits of training across multiple habitats

Deep learning for automated analysis of fish abundance: the benefits of training across multiple habitats

Optical Prior‐Based Underwater Object Detection with Active Imaging

Deteksi Ikan Dengan Menggunakan Algoritma Histogram of Oriented Gradients

Contact Info

Product

Resources

About