Algorithm Product Family

产品简介

Full-scenario Target Detection and Tracking Algorithm Set relies on deep learning visual models, combined with traffic scenario-specific training to enhance robustness in harsh environments. Through multi-dimensional detection and trajectory association algorithms, it completes vehicle and pedestrian recognition, segmentation, and cross-camera tracking. Combined with cloud-edge-device computing collaboration, it achieves high-accuracy, low-latency target perception output, providing a perception foundation for upper-layer services.

产品功能

• Based on years of deep technical accumulation in traffic scenarios, covers harsh roadside environments such as complex lighting, occlusion, large angles, and fast motion
• Supports 2D/3D motor vehicle detection, non-motor vehicle and pedestrian detection and segmentation
• Cooperates with cross-camera target tracking capability to realize efficient collaboration across cloud, edge, and device platforms, providing a stable and precise target perception foundation for upper-layer perception hardware and business platforms.