基于数字图像处理的玉米品种识别研究
详细信息    本馆镜像全文|  推荐本文 |  |   获取CNKI官网全文
摘要
随着信息技术的发展,图像处理与识别技术在种子资源品质检测方面发挥着越来越重要的作用。本文以计算机图像处理技术为主要技术手段,综合运用数字图像、摄影、光学、植物学、色度学、模式识别等方面的知识,对玉米品种的自动识别问题进行了研究。
     一、具体研究内容
     1.玉米种子图像的采集
     为了提高自动识别的准确性同时尽量减少对图像采集的约束条件,对图像拍摄方法进行了研究。选取本地区常用的郑单958、辽单565、京科25作为被识别对象,对不同条件下拍摄的图像进行了分析对比。
     2.玉米种子图像预处理
     背景分割:用几种不同的分割方法进行试验,通过对比验证,采用迭代法进行背景分割。去除噪声:背景分割后图像中存在孤立的噪声点,通过查阅资料,采用自适应中值滤波法。颜色校正:在研究中找到了一种基于标准白板的颜色校正方法。此法根据标准白板的颜色来校正种子图像的颜色,大大提高了图像中种子颜色的准确性。种子分离:主要采用区域标记法来实现种子图像中的标准白板的提取与单粒种子分离。
     3.玉米种子特征提取
     除了利用玉米种子的颜色、形状、大小等常见特征外,本文发现玉米种子白色部分(胚部)与黄色部分(冠部)的面积比例以及形状对玉米种子的品种识别具有一定作用,并对黄白面积比例进行了深入研究,提取了白色部分的面积、黄色部分的面积、黄色部分的颜色等新特征。颜色特征的提取采用RGB模型与HSV模型相结合的方法。
     4.特征集优化与新特征的价值验证
     本文采用基于遗传算法与支持向量机特征优化方法,把新特征与常规特征有区别的进行优化。从两组最优特征子集的识别率与算法运行过程中的特征选择率两方面证实了部分新特征是很有价值。对特征数据进行分析,进一步证实白色部分与黄色部分的面积比例特征的可用性;通过试验验证,黄色部分的蓝色分量与饱和度对识别的价值要优于整个玉米种子的相应特征。
     5.玉米品种识别
     采用支持向量机方法对玉米品种进行识别。三个玉米品种的识别率分别为90.1%、93.5%、96%。
     二、本文的独特之处
     1.提出了一种简单实用的颜色校正方法,使得种子图像的拍摄条件无需固定照明,把玉米品种识别摆脱试验室限制的进程向前推进了一步。
     2.提出了玉米种子白色部分与黄色部分的面积比例等特征作为识别玉米品种的新特征。
Along with the development of information technique, digital image processing and identifying technology becomes a vital one in inspecting the quality of seed by computer. The main purpose of this paper is to deeply research the corn species identification mainly by computer image processing technology. As well this paper involves in other knowledges such as photograph, optics, plant physiology, chroma, pattern recognition etc.
     First, the main contents of the paper are as follows:
     1. The image of corn seed collection
     The method of shooting image is researched on to advance the identification accuracy by computer and reduce restrictive conditions when collect image. Zhengdan958, Liaodan565, Jingke25 which are in common use in this region are selected as the researched objects. This paper analyzes their images which are obtained under different conditions.
     2. The image of corn seed pre-process
     Parttioning Background: This paper decides to adopt iterative method through starting experiment by a few partition methods and comparing the experiment results.
     Smoothing image: After parttioning background, there are same isolated noise points in the image. By consulting datum, self-adjusing median filter is adopted to smoothing image.
     Correcting color: A method is found out in the research that correct image color according to the color of the standard whiteboard. Using the method to correct color improves the accuracy of corn seed color in the image.
     Seed separation: Standard whiteboard and separate seeds are distilled Mainly by labeling area method.
     3. The characteristic of corn seed pick-up
     In addition to a few general characteristics about color, shape and size. This paper finds out the ratio of white part(embryo) and yellow part(crown) of corn seed and their shapes are useful. Then, by deeply researching in this aspect, new characteristics such as the size of the two parts and the color of the yellow part are picked up. RGB model and HSV model are combined together during picking up the color characteristics.
     4. The character of corn seed optimization and the value of new characteristics validation
     This paper uses support vector machine (SVM) combined with genetic algorithm(GA) to discriminatively optimize new characteristics and general characteristics to validate the value of the new characteristic. The recognition rates of the two best character and the select rates of characteristics approve that some new characteristics are very useful. Through analyzing data of the ratio of white part and yellow part, the value of the characteristic are further confirmed. By experimentalizing, the color characteristics of the yellow area are better than the whole corn seed's to the species identification.
     5. The corn species identification
     Using Support Vector Machine to identify the breed of corn seed, the separate rates are 90.1%, 93.5%, 96%.
     Second, the innovative aspects of this paper
     Images of corn seeds are obtained out of fixed illuminate room and a simple and useful method of color correction is put forward which accelerates the identification process to get away from laboratory. And that some characteristics which are found out by partitioning the corn seed to white part and yellow part are valuable to the identification.
引文
[1] 王玺,张宝石.玉米种子纯度鉴定技术研究进展[J].种子,2002,(1):43-45.
    [2] Gunasekarans S, Cooper T M, et al. Image processing for stress crack in corn kernels[J]. Transactions of the ASAE, 1987, 30(1): 266-271.
    [3] Zayas I, Lai F S, Pomeranz Y. Discrimination between wheat classes and varieties by im age analysis[J]. Cereal Chemistry, 1986, 63(1): 52-56.
    [4] Zayas I,Pormeranz Y,L ai F S. Discrimination of wheat and non wheat components in grain samples by image analysis[J]. Cereal Chemistry, 1989, 66(3):233-237.
    [5] Liao K, Reid J J, et al. Corn kernel shape identification by machine vision using a neural network classifier[J]. ASAE Peper, 1992.
    [6] Liao K, Paulsen M R, Reid JF, et al. Corn kernel breakage classification by machine vision using a neural network classifier[J]. Trans of the ASAE. 1994, 36(6):1949-1953.
    [7] Ni- B, Paulsen M R, et al. Corn kernel crown shape identification using image process ing[J]. Trans of the ASAE, 1997, 40(3):833-838.
    [8] Ni- B, Paulsen, et al. Size grading of corn kemels with machine vision[J]. Apple nga gric.1998, 14(5):567-571.
    [9] Chtioui Y, Panigrahi S, Backer L F. Rough sets theory as a pattern classification to ol for quality assessment of edible beans[J]. Transactions of the ASAE, 1999, 42 (4):1145-1152.
    [10] 方如明.计算机图像处理与米的品质检测[J].农业工程学报,1992,8(3):104-112.
    [11] 王丰元,周一鸣.种子形状参数检测的计算机图像处理技术[J].农业机械报,1995,26(2):52-57.
    [12] 周红,刘光蓉等.数字图像处理技术在玉米种子轮廓检测中的应用[J].种子.2004,23(9):90~92.
    [13] 宁纪锋,何东建,杨蜀秦.玉米种子的尖端和胚部的计算机视觉识别[J].农业工程学报,2004,20(3).
    [14] 闸建文,陈永艳.基于外部特征的玉米品种计算机识别系统[J].农业机械学报,2004,35(6).
    [15] 管庶安,周红等.基于Zerni矩的米粒形状模糊分类[J].河南工业大学学报(自然科学版).2005,26(1).
    [16] 魏弘博,吕振肃等.图像分割技术纵览[J].2004,16(2):19-24.
    [17] 黄峰茜,陈春晓,姚均营.图像分割方法的研究进展[J].中国医疗器械信息,2006,12(12).
    [18] 章毓晋.图像处理和分析(M).北京:清华大学出版社,1999.
    [19] dmgf.图像分割的一些简单实现[EB/OL].http://www.5inet.net/Develop/Delphi/031134,TuXiangBanGeDeYiSuoJianChanShiXian.aspx,2006-6-25.
    [20] 石美红.一种高方差的高斯噪声滤波新方法[EB/OL].http://www.china-image.cn/lunwen/2200608181514242045617948.html,2005-9.
    [21] 彭宇新.数字图像处理[R].北京:北京大学计算机科学技术研究所.
    [22] 吴小艳.数字图像处理及识别算法研究与实现[M].新疆:新疆大学,2005.
    [23] 萧绪珊.彩色摄影[M].中国人民大学出版社.1986,9-25.
    [24] Healey G, Kondepudy R. Radio metric CCD cameral calibration and noise estimations. IEEE Transactions on Pattern Analysis and Ma chine Intelligence,1996,18,1994,16(3):267-276.
    [25] P.Shatadal,D.S.Jayas.Digital image analysis for software separation and classification of touching grains:Ⅰ.disconnect algorithm,Transactions of the ASAE.1995,Vol:38(2):653-643.
    [26] P.Shatadal,D.S.Jayas.Digital image analysis for software separation and classification of touching grains:Ⅱ.classification Transactions of the ASAE.1995,Vol:38(2):635-643.
    [27] 油料及其清理[EB/OL].http://www.jxgmjwc.com/yzkc/asl/web/%B5%DA%D2%BB%D5%C2%20%D3%CD%C1%CF%BC%B0%C6%E4%C7%E5%C0%ED.htm.
    [28] 玉米的分类[EB/OL].http://www.mdjagri.gov.cn,2004-10-28.
    [29] 史新海.利用籽粒形态鉴定玉米杂交种纯度的方法[J].种子科技,2001,(2):110.
    [30] 张红军,张文强.玉米种子颜色遗传基础与花粉直感现象利用[J].种子科技,1999,(6):25-26.
    [31] 王玉洁.利用玉米杂交种籽粒形态和幼苗芽鞘颜色鉴定种子纯度的几种方法[J].种子,1998,(1):76-78.
    [32] 周芳,史振声,王志斌.玉米杂交当代籽粒性状及营养成分的变化[J].辽宁农业科技,2007(1).
    [33] 凤·兮.谈颜色空间[EB/OL].http://www.photoshopcn.com/article/photoshop/502.html,2006-3-20.
    [34] 高清柳.基于HSV颜色空间的彩色图像与视频水印算法研究[M].北京:北京化工大学,2006.
    [35] 孙涛.基于颜色空间的图像特征提取的研究[D].吉林:吉林大学,2006.
    [36] 陈兵旗,孙明.Visual C++实用图像处理[M].北京:清华大学出版社,2004.
    [37] 吕一波,路迈西.玉米图像几何特征的提取[J].鸡西大学学报,2006,6(4):68-70.
    [38] 欧阳鑫玉,赵楠楠等.图像分割技术的发展[J].鞍山钢铁学院学报,2002,25(5):364-368.
    [39] 李太富,王靖,谭光仪.图像分割方法的研究[EB/OL].http://www.chenguang-design.com/xhdt/lww/6.htm,2003-11-05.
    [40] 赵忠明.数字图像处理导论[M].西安:西北工业大学出版社,1995.
    [41] 饶洁.基于区域特征及边缘信息的彩色血液显微图像分割[D].四川:四川大学,2006.
    [42] 余成波.数字图像处理及MATLAB实现[M].重庆:重庆大学出版社,2003.
    [43] 黄国祥.RGB颜色空间及其应用研究[D].长沙:中南大学,2002,5-20.
    [44] 郭文川,朱新华.机器视觉技术在谷物识别与分级中的研究进展[J].粮食与饲料工业,2002,(6):50-51.
    [45] 王丰元,周一鸣.种子形状参数检测的计算机图像处理技术[J].农业机械学报,1995,26(2):52-56.
    [46] 边肇祺,张学工.模式识别[M].第二版,北京:清华大学出版社,2000.
    [47] 朱志刚,林学阎.数字图像处理[M].电子工业出版社,1998.
    [48] 毛罕平,徐贵力,李萍萍.基于遗传算法的蔬菜缺素叶片图像特征选择研究[J].江苏大学学报(自然科学版),2003,24(2).
    [49] 苑希民,李鸿雁等.神经网络和遗传算法在水科学领域的应用[M].北京:中国水利水电出版社,2002.
    [50] 余建坤,张文彬,陆玉昌.遗传算法及应用[J].云南民族学院学报(自然科学版),2002,11(4):193-197.
    [51] 刘立平,牛熠.遗传算法综述[J].东莞理工学院学报,2005,12(3):48-52.
    [52] 任平.遗传算法_综述[J].工程数学学报,1999,16(1):1-8.
    [53] 许可,朗兵.遗传算法及其在电力系统中的应用[J].济宁师范专科学校学报,2005,26(3):110-112.
    [54] 王竹卿.基于遗传算法的车间调度问题研究[D].大连:大连理工大学,2006,18-29.
    [55] Vapnik V.Statistical learning theory[M].New York:John Wiley & Sons.1998.
    [56] 李晓宇,张新峰等.支持向量机_SVM_的研究进展[J].控制技术,2006,25(5):7-12.
    [57] 马忠宝.基于支持向量机的中文文本分类系统研究[D].武汉:武汉理工大学,2006,35-45.
    [58] 吴涛.核函数的性质—方法及其在障碍检测中的应用[D].长沙:国防科技大学,2003,4-10.
    [59] 孙德山.支持向量机分类与回归方法研究[D].长沙:中南大学,2004,10-15.
    [60] 张莉.支撑矢量机与核方法研究[D].西安:西安电子科技大学,2002,15-19.
    [61] 郑春红.支撑矢量机应用的关键技术研究[D].西安:西安电子科技大学,2005,16-19.
    [62] 李建民,张钹,林福宗.支持向量机训练算法综述[J].清华大学学报(自然科学版),2003,43(1):32-37.
    [63] 王晓丹,王积勤等.支持向量机训练和实现算法综述.计算机工程与应用,2003,(13):75-78.
    [64] 杨雪.支持向量机多类分类方法的研究[D].哈尔滨:哈尔滨工程大学,2006,17-22.
    [65] 文明,方凯等.一种基于SVM的多类判别算法[J].工业仪表与自动化装置,2006,(6):6-8.
    [66] macwader.LIBSVM的简介及使用[EB/OL].http://macwader.blog.hexun.com/8298118_d.html,2007-03-18.
    [67] 杨宏晖,孙进才,袁骏.基于支持向量机和遗传算法的水下目标特征选择算法[J].西北工业大学学报.2005,23(4).
    [68] Goldberg D.Genetic Algorithm sin Search,Optimization&MachineLeaming.Reading:Addis onWesley, 1989.
    [69] 乔立岩,彭喜元,马云彤.基于遗传算法和支持向量机的特征子集选择方法[J],2006,20(1).
    [70] 陈建安,郭大伟等.遗传算法研究综述[J].西安电子科技大学学报,1998,25(3):363-367.

© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号

地址:北京市海淀区学院路29号 邮编:100083

电话:办公室:(+86 10)66554848;文献借阅、咨询服务、科技查新:66554700