A label field fusion using stochastic ICM optimization technique in image segmentation

This paper presents a new, simple and efficient segmentation approach based an ICM Optimization Technique to obtain an accurate segmentation result.ICM means Iterative Conditional Modes Optimization procedure. ICM is one of the Greedy algorithm that performed at the full resolution to get the final result. This method is accurate and also it should be computationally efficient. In Existing method, so many clustering techniques are available. These techniques are also very expensive and also not accurate. In Proposed method, PRand is calculated for the segmentation result. Among that result, the best one is taken as the starting point of the ICM optimization procedure. It improves the clarity of the image. This technique is successfully applied in Berkeley image database, and which is very simple to implement. The experiments reported in this paper illustrate the potential approach compared to the state-of-the-art segmentation methods recently proposed in the literature.