Frequent incidents of drones colliding with power lines and crashing, SingTown Technology introduces innovative AI visual avoidance solutions.

SingTown Technology嘅智能相機系統，基於OpenMV，能夠喺無人機飛行期間實時識別同避障檢測電線同電纜。利用人工智能同機器視覺演算法，系統能夠喺複雜背景中準確識別細微電線目標，提供早期預警，並協助無人機自動調整飛行路徑以防止碰撞，從而提升飛行安全。

In tasks such as drone inspection, surveying, mapping, and transportation, power lines and cables are among the most threatening potential obstacles. Wires are thin, exhibit varied colours, and are often situated in complex environments, making them difficult to accurately identify using traditional radar or infrared sensors.

To address this challenge, SingTown Technology has developed an AI smart camera solution based on the OpenMV platform, integrating machine vision and computer vision technologies to achieve precise detection and obstacle avoidance warnings for wires and cables.

The system captures flight footage in real-time via a high-frame-rate camera, while AI algorithms analyse linear structural features within the images to identify the position, orientation, and relative distance of power lines relative to the drone. Even in environments with strong light, shadows, or intersecting wires, it can quickly distinguish cable outlines. The recognition results can be transmitted in real-time to the flight control system, assisting the drone in automatically adjusting its course and altitude to avoid collisions.

此方案嘅邊緣部署優勢，令其特別適合對實時性同準確性要求高，但安裝條件有限嘅項目，例如電力巡檢、通訊維護、地質測繪同安防監控等無人機作業場景。OpenMV智能相機具備微型化、低功耗同高幀率嘅特點，能夠輕鬆整合到多旋翼或固定翼平台，喺唔增加負載嘅情況下提供強勁嘅視覺感知能力。

SingTown Technology，憑藉其在邊緣人工智能、機器視覺及圖像識別方面的核心專長，賦予無人機「視覺」能力，主動避開電線及纜線，為智能飛行提供安全護航。這項嵌入式視覺技術正推動無人機產業邁向更智能、更可靠的未來。