论文部分内容阅读
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.
In order to effectively detect the foreign bodies with the similar morphology and color to the background, this paper proposes a method for the detection of foreign bodies based on the difference of the near-infrared absorption characteristics between the foreign body and the background.The method analyzes the difference of the absorption characteristics between the foreign body and the background in the near- In order to determine the best range of detection wavelength for distinguishing the background and multiple foreign objects, a near-infrared spectroscopy imaging system was established to convert the near-infrared absorption difference into the difference of the image features between the foreign object and the background in the near-infrared image. The adaptive image enhancement And binary image processing to extract foreign objects from the background.Experimental results show that the features of the foreign images obtained by this method are obvious and the results of the detection are in accordance with the actual ones.The method can effectively detect foreign objects with similar background features.