An Object Detection Algorithm Based on the Cloud Model

Cloud model is an advanced theory for solving uncertainty problems. Taking the uncertainty of images into account, this paper proposes an object detection algorithm based on the cloud model. First, adopt cloud model theory to transform the imagepsilas qualitative model to its quantitative model. Then, use climbing policy to get different level concepts which represent different level objects. At last, compute a certainty degree and determine which concept each pixel belongs to. Experimental results proved that the algorithm can have a better effect in accuracy of detecting objects and it is good at resolving the edges of different objects.