PHOENIX‑AI Rice Intelligent Color Sorter

Under on‑the‑fly imaging mode, the camera accurately tracks every grain of material and captures subtle differences in morphology and color

Based on sophisticated optical design and self‑developed algorithm models, real‑time image distortion correction is implemented to restore the true geometric structure and surface texture of materials, laying a solid and reliable foundation for subsequent accurate judgment.

Imitating the human‑eye‑to‑human‑brain recognition logic, it features strong anti‑interference capability. It accurately extracts multi‑dimensional material features and rapidly completes quality evaluation through comparative learning. Built‑in dedicated NPU chips provide powerful computing power for high‑speed material inspection, enabling accelerated inference and stable output of the color sorter.

Boasting a self‑developed detection model library with more than 30 dedicated models available, it covers sorting standards for multiple materials. Flexible model switching and rapid iteration endow the color sorter with continuously‑evolving brain‑like capabilities.

Featuring intelligent valve control and multi‑channel collaborative operation, the system responds at microsecond‑level speed to reject defective materials at high speed. With sensitive valve opening‑closing action, it adapts to production‑line load variations and capacity fluctuations, consistently maintaining high yield and high‑efficiency sorting performance.