Hierarchical Citrus Canker Recognition Based on HMAX Features
基于HMAX特征的层次式柑桔溃疡病识别方法

A novel hierarchical approach based on HMAX features is proposed for citrus canker recognition. Citrus canker areas have different appearances and strong partial characteristics. HMAX feature is a scale-tolerant and oriental-tolerant complex feature set which is robust in object recognition. AdaBoost algorithm is efficient in boosting weak classifier to strong classifier. By using HMAX features and AdaBoost algorithm, an efficient classifier can be constructed with few samples. And from bottom up process ca...