基于能量图像的目标识别方法

提出了一种基于AdaBoost算法,利用能量图像进行目标分类的算法.该算法通过灰度图像创建能量图像,并将能量图像作为样本进行机器的集中学习,实现对骑车人、行人和车辆目标的分类识别.根据待检测目标的能量分布,构建了能描述目标的典型特征,并以检测区域的能量差异作为其特征值.重点讨论了骑车人和行人的分类识别问题.实验结果表明,所提出方法对本课题中的特定目标具有较好的识别率,并具有较强的鲁棒性.

Method for objects recognition based on energy of images

A method for multi-object recognition based on AdaBoost is proposed.Energy image created by Grayscale image is regarded as samples for machine learning of cyclist,pedestrian and vehicle recognition.The new typical features are build based on object’s energy distribution by regarding regional energy contrast ratio as feature value.A recognition of pedestrian and cyclist.Experimental results show that this method is robust and has good discrimination in multi-object recognition.