基于c語言圖像形狀特征抽出的應(yīng)用

1、Shape Feature Extraction Based on C LanguageOutlineSummaryExperimentPrincipleIntroductionIntroductiona concept in Computer Vision and Image Processing. applications: military, computer sciences, computer aided diagnose, industry and human machine communication etc.features: color, vein, shape etc.Wh

2、at feature extraction is ?What feature extraction is ?OutlineSummaryExperimentPrincipleIntroductionPrincipleStep1. BinarizationStep2. Noise suppressionStep3. Outline extractionStep4. Image blockingStep5. Feature selectionThe process of shape feature extractionOutlineSummaryExperimentPrincipleIntrodu

3、ctionOriginalimageThree fruits are included: orange, apple and banana(target).Original imageBinarizationBinaryzationBy the equation (1) we can achieve to binary the gray-scale image.BinarizationandnoisesuppressNoise suppressionThis page shows the process of noise suppression2345678G(x,y)G(x,y)1G(x,y

4、)=medium(1,2,3,4,5,6,7,8,g(x,y)OutlineextractionThis page shows the result of outline extraction(In my study, Prewitt method is used for outline extraction, and Hilditch method is used for making the outline thinner)OutlineNoise suppressionPrewittmethod(T is the threshold)ImageblockingAfter outline

5、extraction, we divide the whole image into several parts, and mark them.(in this image, we divide the image into 6 parts)Outline21345Blocking6parametersofeachpartspartarealengthcircularityCenter of gravity11828162.020.8750(27,160)22033172.270.8609(78,91)32992216.790.8000(169,51)45841503.430.2896(147

6、,172)54027401.360.3141(195,121)613755.700.5549(226,199)smaller than 0.5 !Featureselection12345By using some parameters of each parts, we can pick up the target parts what we need(I choose circularity (threshold=0.5) of each parts as the judgment for selecting)Blockingresult6OutlineSummaryExperimentPrincipleIntroductionSummaryl Bythismethodwecanachievetopickupthetargetpartsuccessfully.l Buttherearestil