資源描述:
《灰度圖像特征提取算法研究 畢業(yè)論文》由會(huì)員上傳分享,免費(fèi)在線閱讀,更多相關(guān)內(nèi)容在學(xué)術(shù)論文-天天文庫(kù)。
1、灰度圖像特征提取算法研究摘要圖像特征提取一直是圖像研究和計(jì)算機(jī)識(shí)別中一個(gè)值得探討的問(wèn)題,在計(jì)算機(jī)科學(xué)、醫(yī)療輔助診斷、軍事、工業(yè)測(cè)量等各種領(lǐng)域都正在發(fā)揮著越來(lái)越重要的作用。尤其是醫(yī)療檢索系統(tǒng)中,如何準(zhǔn)確定位和提取關(guān)鍵特征往往是首要解決的問(wèn)題之一,是提高醫(yī)療儀器識(shí)別率的關(guān)鍵。在醫(yī)療檢索系統(tǒng)中準(zhǔn)確的定位和提取圖像的邊緣特征,能夠使醫(yī)生快速準(zhǔn)確的判別出病因。本文針對(duì)邊緣特征提取算法和灰度直方圖特征提取算法進(jìn)行了特征提取試驗(yàn)與分析。邊緣特征提取,本文通過(guò)采取Robets算法,Sobel算法,Canny算法,小波變換算法對(duì)一幅圖像做了不同的邊緣特征提取。結(jié)
2、果表明,本文采取的Canny邊緣特征提取算子法明顯優(yōu)于其他邊緣特征提取算法,且能夠最為接近的描述灰度圖像的邊緣特征信息?;叶戎狈綀D特征,對(duì)于邊緣算子法提取出的灰度邊緣特征圖像,通過(guò)人眼觀測(cè)有時(shí)可能會(huì)觀測(cè)不到明顯的區(qū)別。本文通過(guò)分析其灰度灰度直方圖特征,提取其中的灰度特征量(灰度均值(Mean)、方差(Variance)、熵(Entropy)、能量(Energy)),根據(jù)特征量的變化情況分析圖像邊緣提取效果的差異。本算法編程采用C#語(yǔ)言,在Visualstudio2010平臺(tái)下實(shí)現(xiàn),通過(guò)各種算法對(duì)圖像特征提取的描述實(shí)現(xiàn)其不同的提取效果。結(jié)果顯示C
3、#語(yǔ)言能更好的表示各種不同算法在圖像處理中的編程。關(guān)鍵詞:灰度圖像;特征提??;邊緣特征;灰度直方圖;特征量31灰度圖像特征提取算法研究AbstractResearchandcomputerimagefeatureextractionhasbeenaproblemthatworthexploringinvisualidentification,Incomputerscience,medicaldiagnosis,military,industrymeasurementandsoonvariousfieldsareplayinganincreasi
4、nglyimportantrole.especiallyinmedicalretrievalsystem,howtaccuratelylocateandextractthekeyfeaturesareoftenthefirsttosolveoneoftheproblems,Isthekeytoimprovemedicalinstrumentrecognitionrate.Inmedicalretrievalsystemaccuratelylocateandextracttheimageedgefeatures,Canmakethedoctorq
5、uicklyaccuratelyidentifythecause.Basedonedgefeatureextractionalgorithmandthealgorithmforcalculatinggrayscalehistogramfeatureextractionfeatureextractionexperimentandanalyzed.Theedgefeatureextraction,Algorithminthispaper,byadoptingRobets,Sobelalgorithm,theCannyalgorithm,thewav
6、elettransformalgorithmtoanimagemadeofdifferentedgefeatureextraction.ResultsshowthattheadoptedCannyedgeoperatoroffeatureextractionmethodissuperiortootheredgefeatureextractionalgorithms,Andcandescribethemostclosetotheedgeofgrayimagefeatureinformation.Graylevelhistogramfeature,
7、forgrayleveledgeoperatormethodtoextracttheedgecharacterimages,bythehumaneyeobservationmaysometimesobservednosignificantdifferencebetweenthetwo.Inthisarticle,throughanalyzingthecharacteristicofthegraylevelhistogramgray-scaleand,Extractthegrayscalecharacteristics(grayscaleaver
8、age(mean)andvariance(variance),entropy(entropy),energy(energy),accordingtot