BEIJING, June 27 (Xinhua) -- At first glance, they look like ordinary glasses until they are worn, when cutting-edge agricultural expertise brings the field to life.

These high-tech lenses can scan crops for disease, track fruit growth in real time, and even deliver step-by-step farming instructions directly to the wearer.

These smart farming glasses are currently under development by researchers from the artificial intelligence (AI) research center at China Agricultural University (CAU).

By combining computer vision, deep learning, the Internet of Things and augmented reality, these glasses aim to tackle some of agriculture's toughest challenges, from labor shortages and inconsistent fieldwork to the battle against pests and diseases.

"Traditional methods like manual inspection or fixed cameras cost a lot and are not very efficient," said Li Xiang, the project's lead researcher and an associate professor at CAU. "After extensive field research and technical evaluation, we chose smart glasses, a wearable AI terminal capable of offline operation and real-time feedback, as the next-generation tool for farm supervision and crop monitoring."

The glasses track every move and provide real-time voice prompts and on-screen alerts to guide farmers with instructions like, "Please prune moderately here" or "This fruit is ready for harvesting."

According to Li, the glasses are empowered by a suite of AI models, from object detection and hand tracking to motion recognition and phenotype analysis. These allow the system to recognize specific targets like leaves, fruits, or pest damage even in complex environments. When anomalies are detected, the glasses issue instant audio or visual alerts to prompt quick intervention and minimize crop loss.

During its technical tests in a grape farm in southwest China's Yunnan Province, the technology has cut pruning errors from 10 percent to less than 2 percent, saving farmers significant losses.

"Right now, we're focusing on high-value crops such as grapes, strawberries and tomatoes," Li said. "Not all produce suits this method. We target technically demanding cash crops, including citrus, blueberries, apples and pears."

Developing field-ready AI glasses was no easy task. The team had to compress large AI models to run efficiently on portable devices, employing technologies such as quantum neural networks to enable offline operation in remote areas with limited internet access.

"The development of smart glasses is not just a technical challenge," Li said. "It requires breakthroughs in three areas -- data collection, hardware design and AI optimization. Through large-scale field trials, precise hardware selection and iterative algorithm refinement, we're steadily overcoming these hurdles."

The team also has plans to develop personalized versions for urban gardening. "Future models might be lighter, voice-interactive, or equipped with virtual displays to assist with balcony farming," he said. "They could help people spot over-fertilization, underwatering, or early-stage pests, creating a potential niche market for hobbyists."

"What we're building is the idea that AI models will become as fundamental to farming as seeds or fertilizer," Li said. "Agriculture will move beyond relying on good weather or gut instinct. It will be a smart industry capable of autonomous judgment, rapid response and continuous optimization."

In this transformation, he said, AI will serve as agriculture's "new brain" by not just identifying problems, but offering solutions and automating decisions across production, management and services.

"AI smart glasses are just one node in our broader smart farming ecosystem," Li said. "To truly achieve intelligent agriculture, we must complete a closed-loop system encompassing data collection, intelligent analysis, precise execution and continuous refinement."

"Our team will continue to strive to make that vision a reality," he added. ■