摘要
烟包解包在香烟自动化生产中是一项关键的技术,它对解包的每个过程要求很高的精确度和实时性。论文采用固定的Basler摄像头获取烟包状态的原始图像,充分利用彩色图像的信息,提出了一种在不同的颜色空间中基于物料颜色特征的识别算法。该算法能针对烟包不同状态,通过对RGB三通道分析、滤波、边缘提取,能够识别出带有纸板的烟包与去除塑料薄膜烟包的两种状态;基于HSV模型,通过对物料不同状态的颜色矩分析,能够区分出带有塑料袋的两种不同状态,然后对这两种状态采用图像分块的理论,利用RGB的颜色特征,能够识别这两种不同的状态。该算法实现了对烟包四种状态的区分。
This paper studies the original image obtained by means of the static Basler CCD camera.According tothe information of color image,a new image recognition algorithm is proposed based on the color features in different color model.The algorithm can identify the cigarette package with cardboard and the Cigarette package without plastic film through RGB analysis,filtering,edge detection.Based on HSV model,the cigarette package with plastic film can be distinguished by analyzing the material color moments of the different states;then the method combines image block theory and uses the RGB features,which is able to identify the two different states.
出处
《工业控制计算机》
2013年第7期63-64,66,共3页
Industrial Control Computer
关键词
颜色特征
物料识别
烟包解包
边缘提取
图像分块
color characteristics
material recognition
cigarette package unpacking
edge detection
image block