基于近红外波长成像的异物检测新方法(英文)

Detection of Foreign Materials Based on Near Infrared Wavelength Imaging

为有效检测出与背景形态、颜色极其相似的异物,根据异物与背景近红外吸收特性的差别,提出近红外光谱成像检测异物的方法.该方法分析近红外波段中异物与背景吸收特性差异随波长变化的规律,确定了区分背景与多种异物的最佳检测波段范围,建立近红外光谱成像系统,将近红外吸收特性差别转化为近红外图像中异物与背景的图像特征差别,利用自适应图像增强和二值化图像处理从背景中提取异物.实验结果表明,该方法获取的异物图像特征明显,检测结果与实际相符,此方法可有效检测与背景特征相似的异物.

For effective detection of foreign fibers having almost the same shape and color as the background, an infrared spectral imaging method based on the absorption characteristic discrimination between foreign materials and background is proposed. According to the method, the relation between the absorption characteristic discrimination and the wavelength in the region of near infrared （NIR） is analyzed, and the optimal band for discriminating several types of foreign materials from background is determined. The absorption characteristic of foreign materials is transformed into the image feature in the developed infrared spectral imaging system. Furthermore, the foreign materials are extracted from the background using adaptive image enhancement and binary image processing. The experimental result indicates that the image features of foreign materials are obvious and the conclusion is consistent with the fact. And it provides an effective method to detect foreign materials from background.