Computer Counts Seals, Saves Scientists Time

Spread the love

Computer counts seals from aerial photographs with lightning speed and reliability. Based on their spatial patterns, the tiny dots on the aerial images can even be assigned to one of the two major species of seals in the Wadden Sea. That is shown in the thesis that marine biologist Jeroen Hoekendijk will defend on January 26 in Wageningen.

Computer-Aided Seal Counting: A Technological Breakthrough in Marine Mammal Monitoring

Related Video

Published on: July 24, 2022 Description: A look into how CSGRT use photo identification to monitor and protect grey seals. Filmed by Muddy Duck Productions ...

Every Seal Counts

In the vast expanse of the world’s oceans, marine mammals like seals play a crucial role in maintaining the delicate balance of marine ecosystems. However, with the looming threat of climate change and the disappearance of sea ice, these majestic creatures face an uncertain future. To better understand and protect these vulnerable species, scientists are turning to cutting-edge technology for assistance.

Enter Jeroen Hoekendijk, a marine biologist who has harnessed the power of computers to revolutionize the way we count seals. Through his innovative research, Hoekendijk has demonstrated that computers can not only count seals from aerial photographs with remarkable speed and accuracy but also distinguish between the two major seal species found in the Wadden Sea: harbor seals and gray seals.

Computer-Aided Seal Identification and Counting: A New Era in Marine Mammal Monitoring

Hoekendijk’s groundbreaking work has opened up new possibilities for monitoring marine mammal populations. By leveraging artificial intelligence (AI), computers can analyze vast amounts of aerial imagery, identifying and counting seals with a level of precision that surpasses human capabilities. This technological breakthrough has the potential to transform the way scientists study and conserve marine mammals.

Training Computers to Recognize and Count Seals

To train computers to recognize and count seals, Hoekendijk employed a technique known as “deep learning.” This involves presenting the computer with a large collection of aerial photographs containing seals, along with verified counts of seals in each image. The computer then learns to identify seals based on their shape, size, and spatial distribution.

Distinguishing Between Seal Species: Harnessing AI for Species-Level Discrimination

One of the remarkable aspects of Hoekendijk’s research is the ability of computers to differentiate between harbor seals and gray seals. These two species exhibit distinct spatial patterns when resting on sandbanks, with harbor seals being more dispersed than gray seals. By analyzing these spatial patterns, computers can accurately assign each seal to its respective species, even in images where the resolution is too low to distinguish between them based on external characteristics.

The Wadden Sea: A Perfect Field Lab for Computer-Aided Seal Counting

The Wadden Sea, a vast intertidal area located in the North Sea, has served as an ideal testing ground for Hoekendijk’s research. The Wadden Sea is home to both harbor seals and gray seals, and its well-documented seal populations provide a wealth of data for training and validating computer models.

Expanding the Scope of Marine Mammal Monitoring: From the Wadden Sea to the Arctic

The success of Hoekendijk’s research in the Wadden Sea has paved the way for its application in other regions, including the Arctic. The Arctic is home to a diverse array of marine mammals, including harp seals, whose populations are particularly vulnerable to the effects of climate change. By utilizing computer-aided seal counting techniques, scientists can monitor harp seal populations across vast expanses of the Arctic, providing valuable insights into their status and conservation needs.

Satellite Imagery: The Next Frontier in Computer-Aided Seal Counting

Hoekendijk envisions a future where satellite imagery will play an increasingly significant role in marine mammal monitoring. As the resolution of satellite images continues to improve, it will become possible to identify individual seals from space. Combined with the power of deep learning, satellite imagery could revolutionize our ability to track and protect marine mammals worldwide.

Wrapping Up: Computer-Aided Seal Counting – A Game-Changer in Marine Mammal Conservation

Hoekendijk’s research represents a significant leap forward in the field of marine mammal monitoring. By harnessing the power of computers and artificial intelligence, scientists can now count seals with unprecedented speed, accuracy, and species-level discrimination. This technological breakthrough has the potential to transform the way we study and conserve these iconic marine creatures, helping us to better understand and protect them in a rapidly changing world..

FAQ’s

1. What inspired Jeroen Hoekendijk to develop a computer-aided seal counting system?

Hoekendijk’s inspiration came from the need to monitor seal populations more efficiently and accurately in the face of climate change and the disappearance of sea ice.

2. How does the computer-aided seal counting system work?

The system utilizes artificial intelligence (AI) and deep learning techniques to analyze aerial photographs, identifying and counting seals based on their shape, size, and spatial distribution.

3. Can the computer-aided seal counting system distinguish between different seal species?

Yes, the system can differentiate between harbor seals and gray seals by analyzing their spatial patterns when resting on sandbanks.

4. Why is the Wadden Sea a suitable location for testing the computer-aided seal counting system?

The Wadden Sea is home to both harbor seals and gray seals, and its well-documented seal populations provide a wealth of data for training and validating computer models.

5. How can the computer-aided seal counting system be applied in other regions, such as the Arctic?

The system can be applied in other regions by utilizing satellite imagery and adapting the deep learning models to the specific characteristics of the target seal species and their environment.

Links to additional Resources:

1. www.wur.nl 2. www.wageningenur.nl 3. www.researchgate.net.