Sobel算子在衣物纹理类型检测中的应用研究

Research on Application of Sobel Operator for Clothing Texture Type Detection

衣物纹理是指衣物表面的纹络现象。在越来越发达的智能图像处理过程中,针对衣物纹理的检测也变得愈发平常,人们期望能够使用计算机自动识别衣物表面的纹理类型,如横竖条纹、方格及碎花等,从而进行基于纹理类型的相应操作。文中提出了一种利用Sobel算子进行梯度方向统计来实现衣物纹理类型检测的方法,其中对各纹理类型的判定依据来源于实验分析过程,经测试表明,该方法及相应判定依据能够有效地检测出衣物常见的纹理类型。

Clothing texture is the phenomenon of surface veins.It is more and more common for the detection of significant texture of clothing in the developed intelligent image processing.People also expects to use the computer automatically identify the surface texture of clothing types,such as stripes,squares,floral and so on,and thus for the corresponding operation based on texture types.It provides an implementation using Sobel operators for gradient direction statistics to detect the clothing texture types,which are based on experimental analysis process.The experiment proved that this method can effectively detect common types of clothing and textures.