水果图像的背景分割和边缘检测技术研究_应义斌.pdfVIP

浙江大学学报()　26( 1): 35～ 38, 2000 Journal of Zhejiang University ( Agric. Life Sci. ) : 1008-9209(2000) 01-0035-04 应 义 斌 ( , 310029) 　: 根据区域阈值法分别对所采集的彩色图像的红、绿、蓝三帧图像进行背景分割, 现在 B分量 灰度直方图中利用双峰法选择阈值进行背景分割的效果最好;利用多种微分算子提取水果图像的阶跃 性边缘,并用 Hilditc 细线化方法对已提取边缘的图像进行了细线化处理,获得了较好的边缘提取和细 化处理效果;所研究的背景分割和边缘检测技术的处理效果可以满足进一步进行水果的尺寸、果梗、形 状和表面缺陷等检测的要求. 　　: 水果; 形状; 背景分割; 边缘提取; 细化; 图像; 机器视觉 : TP391. 41; TP391. 76　　　: A - ( . . . . , , 310029, ) YIN G Yi bin Coll of Agric Eng Food Sci Zhejiang University Hangzhou China Study on background segment and edge detection of fruit image using machine vision . Journal of Z ejiang University ( Agric. Life Sci. ) , 2000, 26( 1): 35～ 38 Abstract: In t e lig t of t e c aracteristics of gray level variations in t e image, bimodal gray region seg- ( , , ment m et od was adopted to segment t e pear from t e background in t e t ree images red green blue) . T e best segment result was appeared in B gray level istogram. Four different edge operator w ere used to extract t e step-like edge of t e fruit image, and t e Hilditc t inning met od was used to fulfil edge t inning. As t e results s own, t e adopted met od of background segment and edge detection could , , , . satisfy t e requirem ent of detecting size s ape stem and surface defect of f ruit Key words: fruits; s ape; background segment; edge extraction; t inning; image; mac ine vision 　　, . , ., 、、 . , ,、, 1　图像与背景的分割 ,,