A drone slowly ascended into the sky above the mangrove forest in Zhanjiang city, Guangdong province. Subsequently, orange boxes flashed on the screen one by one.

Biological intruders, including sonneratia and smooth cordgrass, had been accurately targeted, and the system had automatically recorded their location and quantity.

The machine was operated by relevant departments in Zhanjiang, a coastal city in Guangdong province. It uses a recognition algorithm model to guard the regional forests.

According to a statement released by the Zhanjiang city government on Thursday, the algorithm was developed by a local sci-tech company recently registered on the Canton Data Exchange.

The protection of mangrove forests in Zhanjiang has entered the "algorithm era", with faster identification, earlier warnings, and more accurate governance. AI is helping relevant departments and managers respond more scientifically to the challenge of invasive species proliferation.

Lin Dong, president of Zhanjiang's UAV application industry association, said invasive species are affecting the local ecosystem in Zhanjiang. "Taking smooth cordgrass as an example, its tip has inverted spikes like dogtail grass, which is an important feature for artificial identification," he said. "During high tide, it only appears slightly above the water surface and is difficult to detect with the human eye."

He added that the algorithm has opened up a path for "aerial identification of foreign biological invaders, playing an important role in guarding the city's sea forest".