{"id":5707,"date":"2023-03-20T16:48:35","date_gmt":"2023-03-20T16:48:35","guid":{"rendered":"https:\/\/dsrc.haifa.ac.il\/?p=5707"},"modified":"2023-09-19T08:04:27","modified_gmt":"2023-09-19T08:04:27","slug":"3d-imaging-for-coral-reef-ecology","status":"publish","type":"post","link":"https:\/\/dsrc.haifa.ac.il\/?p=5707","title":{"rendered":"3D Imaging for Coral Reef Ecology<\/strong>"},"content":{"rendered":"\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t
\n\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t

3D Imaging for Coral Reef Ecology\n<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/section>\n\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t
\n\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t

Matan Yuval# - Department of Marine Sciences
Naama Pearl#, Amit Peleg, Dan Tchernov, Tali Treibitz - Department of Marine Technologies
Avi Bar Massada - Department of Biology and Environment<\/h2>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/section>\n\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t
\n\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t
\n\t\t\t\n\t\t\t\t\t\t<\/span>\n\t\t<\/div>\n\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t
\n\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t

\u00a0Machine Learning
Deep learning
Neural Networks
Computer vision
Image processing
<\/strong><\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t

\n\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t

Biodiversity<\/strong><\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t

\n\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t

Seed Grant 2021<\/strong><\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/section>\n\t\t\t\t

\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t

Coral reefs are some of the most important ecosystems on the planet. Yet, they are extremely challenging to study across spatial and temporal scales because they are intricate, diverse, and remote. Over the past decade, photogrammetry has emerged as a popular and cost-effective method for coral reef surveys and 3-Dimensional (3D) mapping. However, much work remains on extracting the ecological information concealed in the 3D models such as automated taxonomic classification and structural complexity assessment. To bridge these gaps, we leverage computer vision and deep learning in various tasks, from reef-scale 3D change-detection to colony-scale studies.\u00a0<\/span><\/p>

In our recent study, we employed geometrical measurements to reveal the <\/span>response of coral reefs to a powerful storm<\/a><\/strong><\/span> [1]. We used novel computational tools and detailed 3D reconstructions captured at three time-points over three years. Our data-set <\/span>Reefs4D <\/span><\/i>of 21 co-registered models enabled us to calculate the differences at seven sites over time. Our novel analysis framework is widely transferable and useful for research, monitoring, and management (Fig. 1 A, B).\u00a0<\/span><\/p>

At the moment, we are working on 3D instance segmentation of single coral colonies. We use deep-learning to classify the coral\u2019s segments by their developmental stage. Segmenting the building blocks of corals: Zooids, enables us to study coral growth patterns and their relation to the environment\u2019s structure. This is an important aspect of coral reef ecology that has rarely been studied before, especially due to the large amount of zooids in a colony. For example, the biggest colony in our study contains over 3000 segments (Zooids)(Fig. 1 C). Incorporating AI in our workflow enables us to generate a large dataset with minimal intervention, and to revisit classic studies as well as explore new ideas regarding coral colony formation.<\/span><\/p>

Altogether, our work can help to understand coral reef ecosystems, and how to manage and protect them in light of climate-change and their global decline.<\/span><\/p>

\"\"<\/p>

Figure 1: A) 3D models of a coral reef in 2019, 2020, and 2022 (left to right). The highlights on the left show the differences.\u00a0\u00a0<\/b><\/p>

  1. B) Box counting and Alpha shapes for fractal dimension calculation (see our <\/b>preprint<\/b><\/a><\/span>). C) 3D instance segmentation of coral zooids in two coral genera for studying coral colony formation (in prep.).\u00a0\u00a0<\/b><\/li><\/ol>

    View more 3D models at:\u00a0 <\/span>https:\/\/sketchfab.com\/Marine_Imaging_Lab<\/a>\u00a0\u00a0<\/strong><\/span><\/p>

    1) Yuval, Matan, et al. “3D Imaging Reveals Resilience: Storm Impact on Coral Reef Structural Complexity.” <\/span>bioRxiv<\/span><\/i> (2022): 2022-12<\/span><\/p>\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/section>\n\t\t\t\t

    \n\t\t\t\t\t\t
    \n\t\t\t\t\t
    \n\t\t\t
    \n\t\t\t\t\t\t
    \n\t\t\t\t
    \n\t\t\t\t\t\t\t
    \n\t\t\t
    <\/div>\n\t\t<\/div>\n\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/section>\n\t\t\t\t<\/div>\n\t\t","protected":false},"excerpt":{"rendered":"

    Matan Yuval, Naama Pearl, Amit Peleg, Dan Tchernov, Tali Treibitz, Avi Bar Massada<\/p>\n","protected":false},"author":93,"featured_media":7024,"comment_status":"open","ping_status":"open","sticky":false,"template":"elementor_header_footer","format":"standard","meta":{"_acf_changed":false,"_eb_attr":"","footnotes":""},"categories":[86,31,60,63,62,64,66,23,81],"tags":[33,56,57,34,58],"class_list":["post-5707","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-86","category-biodiversity","category-computer-vision","category-deep-learning","category-image-processing","category-machine-learning","category-neural-networks","category-research-blog","category-seed","tag-computer-vision","tag-deep-learning","tag-image-processing","tag-machine-learning","tag-neural-networks","entry","has-media"],"acf":[],"aioseo_notices":[],"aioseo_head":"\n\t\t\n\t\n\t\n\t\n\t\n\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t