首页|基于机器视觉的烤房成杆鲜烟叶成熟度判别研究

基于机器视觉的烤房成杆鲜烟叶成熟度判别研究

扫码查看
鲜烟叶成熟度判别是生产优质烟叶的关键环节之一,也直接影响后续烟叶环节的烘烤质量.基于机器视觉方法,使用StackingClassifier集成学习策略,建立了成杆鲜烟叶成熟度判别模型,使用外测样本进行两个批次的成杆鲜烟叶成熟度等级判别.结果表明,所建立的模型在两个批次样品中的判别准确率均在94%以上;在成杆鲜烟叶主体成熟度档次比例的预测上,预测比例与真实比例的平均绝对误差为6%.该方法实现了批量化的烟叶成熟度预测,为烤房烟叶成熟度的精准把控提供了参考.
Research on maturity recognition of fresh tobacco leaf bundles in baking chambers based on machine vision
The identification of tobacco leaf maturity is one of the key links to produce high-quality tobacco leaves,and also directly affects the curing quality of subsequent tobacco leaves.Based on machine vision method and StackingClassifier integrated learning strategy,the maturity discrimination model of fresh tobacco leaves was established,and the maturity level of two batches of fresh tobacco leaves was determined by external test samples.The results showed that the accuracy of the model was above 94%in two batches of samples.The average absolute error between the predicted proportion and the real proportion was 6%.This method realized the prediction of the maturity of bulk tobacco leaves and provide a reference for the accurate control of the maturity of tobacco leaves in baking chambers.

flue-cured tobaccomaturity recognitionbaking chambersmachine visionintelligent baking

喻曦、李洪明、戴恩、孙五三、朱法亮、赵文军、李跃平、王家绪、胡慧新

展开 >

云南省烟草公司昆明市公司,云南 昆明 650051

上海创和亿电子科技发展有限公司,上海 200082

烤烟 成熟度识别 烤房 机器视觉 智能化烘烤

2024

安徽农学通报
安徽省农学会

安徽农学通报

影响因子:0.275
ISSN:1007-7731
年,卷(期):2024.30(24)