基于视觉显著性的大米垩白区域分割方法  

A Segmentation Method for Chalky Region of Rice Based on Visual Saliency

作　　者：车梦凡 CHE Mengfan(Zhejiang Jinhua Technology&Trade Polytechnic,Jinhua Zhejiang 321000,China)

机构地区：[1]浙江金华科贸职业技术学院,浙江金华321000

出　　处：《信息与电脑》2024年第22期70-72,共3页Information & Computer

基　　金：浙江省教育厅一般科研项目(项目编号:Y202352948)。

摘　　要：大米的垩白是影响大米品质的因素之一,随着农业智能化的发展,机器视觉在大米垩白区域分割中的方法层出不穷。针对大米垩白区域分割方法只从图像的空域设计,分割后的垩白区域缺少细节部分,本文从空域和频域出发,设计了一种提取大米垩白区域的视觉显著性分割方法。该方法可以有效分割出大米的垩白区域并能保留垩白区域的细节。实验结果表明,本文研究算法的Dice值达到了96.32%。Chalkiness in rice is one of the factors affecting its quality.With the development of agricultural intelligence,machine vision has produced a myriad of methods for segmenting chalky regions in rice.As existing methods for segmenting chalky regions in rice often focus solely on the spatial domain of images,the segmented chalky regions often lack detailed parts.This paper,designs a visual saliency segmentation method for extracting the chalky region of rice from the spatial domain and frequency domain.This method can effectively segment the chalky regions of rice while preserving the details of these regions.The experimental results show that the Dice coefficient of the research algorithm presented in this paper has reached 96.32%.

关 键 词：视觉显著性 图像分割 垩白区域 

分 类 号：TS2[轻工技术与工程—食品科学与工程]