摘要： 在基于计算机视觉技术对无土栽培番茄营养元素缺乏智能识别研究中,如何选择 出对缺素叶片分类能力强的特征项组合是识别诊断面临的关键问题,文章利用遗传算法对提 取的缺素叶片图像众多颜色和纹理特征项进行优化选择,以达到诊断识别用的信息最优,实 验表明,经过优化的特征组合明显优于人工选择的特征组合分类能力?

关键词: 特征项的选择, 遗传算法, 番茄, 叶片, 图像

Abstract: In the research of diagnosing tomato disease of nutrient deficiency intellectively in the soiless agriculture, feature selection is a key question in pattern recognition and affects the design and performance of the classifier. In this paper, using genetic algorithm (GA) to select features of leaves image is studied in order to get best information for diagnosing. The experimental res ults show that the effectiveness of features selected by GA is much better than the effectiveness of features selected by human.

中图分类号: 

• TP391.41