CAJ | 학술논문

　　针对棉花机械打顶存在的问题，运用机器视觉技术通过对棉株株顶的精确识别实现精确打顶。对采集的棉株图像利用MATLABR2011a 软件进行棉株图像处理，分割出棉株株顶和叶片等其它部分；提取棉株株顶和叶片等其它部分的R，G，B等3通道的均值、方差与标准差，作为棉株株顶区别于叶片等其它部分的特征；最后，利用BP 神经网络对棉株株顶进行识别。结果表明，颜色特征能够很好地区分出棉株株顶和其它部分。
　　침대면화궤계타정존재적문제，운용궤기시각기술통과대면주주정적정학식별실현정학타정。대채집적면주도상이용MATLABR2011a 연건진행면주도상처리，분할출면주주정화협편등기타부분；제취면주주정화협편등기타부분적R，G，B등3통도적균치、방차여표준차，작위면주주정구별우협편등기타부분적특정；최후，이용BP 신경망락대면주주정진행식별。결과표명，안색특정능구흔호지구분출면주주정화기타부분。
Extracting cotton top features as cotton features for cotton identification , in order to solve the problem which cotton toping by machine meted, the Machine Vision technology was explored in this paper .MATLABR2011a software was used to the segmentation of the cotton top features .Leaves and other top parts were extracted in this paper , the means, variances and standard deviations of the R ,G,B channel were used as features to distinguish the cotton top and the other parts.Finally, the BP neural network was used to the identification of the cotton top in this paper .Wide data-base of images were used to test the results and the results show good accuracy in the identification of cotton top .