论文部分内容阅读
基于仿生学思想设计一种果实簇识别与采摘顺序规划的新方法。该方法以Itti视觉注意基本模型为基础,通过改进视觉特征整合方法,构建依赖先验知识的视觉显著图,改善果实簇的识别效果;借鉴“赢者取全”的生物神经竞争机制和采摘专业人员的操作经验,设计距离、面积和显著度加权择优的采摘顺序规划策略,提高采摘工作效率。实验结果表明,本文所设计的算法能够识别20余种常见果蔬,识别正确率高于93.36%;果实簇的采摘顺序规划结果符合人工采摘的专家经验。
Designing a New Method of Fruit Cluster Recognition and Picking Sequence Planning Based on Bionics Theory. Based on the Itti visual attention basic model, this method constructs a visual saliency map which depends on prior knowledge to improve the recognition effect of fruit clusters by improving the visual feature integration method. Referring to the “winner-take-all” biological neural competition mechanism and picking professional Personnel operational experience, design distance, area and saliency weighted picking sequence planning strategy to improve the efficiency of picking. The experimental results show that the algorithm designed in this paper can identify more than 20 kinds of common fruits and vegetables, and the recognition accuracy is higher than 93.36%. The picking sequence of fruit clusters is in line with the expert experience of artificial picking.