如何計算圖像中物體空間先驗信息？

  TextonBoost: Joint Appearance, Shape and Context Modeling for Multi-class Object Recognition and Segmentation   Indoor Scene Segmentation using a Structured Light Sensor

>>阅读原文<<