基于高光谱的颜色辨别方法研究

在颜色测量领域区分相似颜色的样品是非常困难的。测量颜色的准确性和高效性对于工业上的应用非常重要。提出了一种基于超光谱成像技术的色彩测量的方法，并设计制成原理样机系统。该系统能够快速准确的测量彩色样品的光谱，并在分析后可得丰富的颜色数据与颜色坐标。该方法克服了传统测色方法“测谱不成像，成像不测谱”的局限性。为了评估系统的性能，进行分析和实验：比较细分的每个波段的信噪比，并使用光谱匹配技术来比较彩色照相机和所设计的系统在颜色测量方面的优缺点。结果表明，本系统提供了一种更精确的颜色测量方法，可以有效地测试产品颜色的质量。

Color Discrimination Based on Hyperspectral Imaging Method

It is very difficult to distinguish samples with similar colors in color measurement. Accuracy and efficiency are very important for industries’ application. A method for color measurement based on hyper- spectral imaging technology was proposed, and a prototype was integrated in this paper. The system is able to measure the spectrum of the colorful samples rapidly. The color analysis or color coordinates were calculated after measurement. This method provided the image and spectrum at the same time. To evaluate the performance of the system, analysis and experiment were also performed. We compared the signal-to-noise ratio of each bands that are subdivided, and used the spectral mapping technique to compare the advantages and disadvantages of the color cameras and the proposed system. The results show that we proposed a more accurate method for color measurement and it can test the quality of products efficiently.