Tencent cooperates with WWF to build a digital recognition platform with artificial intelligence technology to protect snow leopards in Gansu Province, China.

To protect snow leopards, Tencent has cooperated with WWF to build a digital platform to track the tracks of snow leopards and collect various key data to protect snow leopards and their habitats.

adopt AI Recognition technology, the digital platform can identify snow leopards in pictures and videos and screen out empty scenes. This greatly improves the efficiency of species identification from massive monitoring data.

The digital platform can also conduct data analysis and export intuitive results, showing the number, location and activities of snow leopards living in the region. Forest rangers and environmental experts can formulate protection strategies based on the analysis results.

A snow leopard was detected by Tencent digital recognition platform in Qilian Mountain National Park.

Difficulties in the process of conservation

Snow leopards are very rare, and there are only a few left in the world four thousand Only. World Conservation Union List snow leopards as “ vulnerable ” Species, and their total number is still declining.

Global 60% The habitat of snow leopards is located in China, but the survey area of snow leopards in China is less than its habitat 2% 。 Snow leopards are at the top of the food chain, so their living conditions are also an indicator of the health of China's high-altitude ecosystem. It is of great significance to monitor and protect them.

However, in the process of monitoring and protecting snow leopards, forest rangers, researchers and professional conservationists around the world face many difficulties. They need to collect data in complex terrain, identify species in massive pictures and video files, and conduct complex data analysis to develop follow-up protection measures and ecological research.

A forest ranger in Qilian Mountain National Park said: “ Snow leopards live in rocky areas, which are difficult for humans to access. The patterns on their bodies are conducive to camouflage, thus increasing the difficulty of identification. More than half of the photos and videos of snow leopards we took were incomplete, which made it difficult for us to formulate effective protection strategies. ”

Snow leopard habitat shape It is steep and hard for human beings to reach.

The data collected in May this year shows that the accuracy rate of the digital platform in identifying snow leopards has reached 98% , can save more than  50% Manpower.

Recently, the project team is developing the function of uploading data to the cloud. The camera set in the reserve can automatically upload pictures and videos to the cloud system. Forest rangers can collect data remotely and obtain results in real time.

In addition to snow leopards, the system can also detect twenty More than species, including red fox, wolf, lynx, brown bear, etc.  

Science and technology are good

This carry AI The digital platform of technology has also been selected two thousand and twenty-one Fifteenth meeting of the Conference of the Parties to the Convention on Biological Diversity（ COP15 ）Biodiversity of 100+ Case.

The project team hopes to extend the system to snow leopard habitats around the world and establish cross-border biodiversity conservation cooperation.

An environmental expert of WWF said: “ We hope to share technology with other countries to jointly protect snow leopards around the world. In the process of protecting snow leopards, we also protected the living environment of other species in the area. ”

Qilian Mountain National Park is one of the largest nature reserves in China, with rich natural species.