How do hyperspectral cameras make color measurements?
In today's era of rapid development of science and technology, color measurement has a vital position in many fields, from product quality control, artistic creation to scientific research. As an advanced optical device, hyperspectral camera brings a new, more accurate and comprehensive solution for color measurement.
一、the basic principle of hyperspectral camera
The working principle of hyperspectral cameras is based on the fine capture of spectral information. Unlike traditional cameras, which can only record the color information of the three channels of red, green and blue, hyperspectral cameras can divide the spectrum into many narrow bands in a wide spectral range such as visible light to near infrared, usually up to hundreds or even more. For example, it can divide the spectral range of 400-1000nm into bands with very small intervals, such as 1nm or smaller intervals. When light shines on the surface of the measured object, the reflection, absorption and transmission characteristics of the object to different wavelengths of light are different. Through its special optical system and detector, the hyperspectral camera collects the intensity of the light signal of each band in turn, so as to construct the spectral reflectance curve of the object. This curve records in detail the reflectivity of objects at various wavelengths and is the basic data source for color measurement.
二、the specific process of color measurement
(1) Calibration
Calibration is a critical step before using a hyperspectral camera for color measurement. The purpose of calibration is to establish an accurate correspondence between the spectral data captured by the camera and the true color values. Standard whiteboards with known spectral properties are often used as calibration references. Standard whiteboards have stable and precisely known reflectance at various wavelengths. The hyperspectral camera takes pictures of the standard whiteboard, records its optical signal intensity in each band, and calculates the response function of the camera according to the known spectral reflectance data of the standard whiteboard, so as to correct the possible spectral deviation, dark current noise and other error factors of the camera, and ensure the accuracy and reliability of the subsequent measurement data.
(2) Image collection
After the calibration is completed, the image of the target object can be acquired. When a hyperspectral camera takes pictures of an object, it obtains the intensity information of the light reflected by the object band by band according to the preset spectral band range and resolution. For example, for each pixel in an image, its reflected light data across multiple spectral bands is recorded. If the camera divides the spectral range into 200 bands, then each pixel will have 200 corresponding spectral reflectance values. Together, these data form a three-dimensional data cube, where the two-dimensional plane represents the spatial position information of the image (x, y coordinates), and the third dimension represents the spectral band information (λ). In this way, the hyperspectral camera not only records the color and appearance information of the object, but also contains its spectral characteristics information, which provides more abundant data than traditional cameras.
(3) Data processing and color calculation
The massive spectral data collected need to go through complex data processing to get the final color measurement results. First of all, the data should be preprocessed, including removing noise, correcting spectral distortion and other operations. Then, the color is calculated according to a specific color model and algorithm. In the field of color science, the commonly used color models are CIE XYZ, CIELAB, etc. Taking the CIELAB color model as an example, it represents color as three coordinate values based on the human eye's perception characteristics of color: L represents the brightness, a represents the red-green degree component, and b * represents the yellow-blue degree component. By combining the spectral reflectance data collected by the hyperspectral camera with the spectral power distribution of the standard illumination body (such as the D65 standard light source), and integrating according to the color matching function, the coordinate value of the object in the CIELAB color space can be calculated, so as to accurately describe the color attribute of the object. Such as color depth, tone and saturation. In addition, color difference can also be calculated by comparing the color coordinate values of different objects or different parts of the same object, which is used to evaluate the consistency or degree of change of color.
三、the advantages of hyperspectral camera color measurement
(1) High precision and high resolution
Hyperspectral cameras provide extremely high spectral resolution, which allows them to capture extremely fine color differences in color measurements. For example, in some industries that require very high color accuracy, such as high-end printing, cosmetics production, etc., it can accurately distinguish color changes that are difficult for the human eye to detect, ensuring the consistency of product color and high quality standards. Its high-precision measurement results help to improve the quality control level of products and reduce the rate of defective products caused by color deviation.
(2) Rich spectral information
In addition to the tristimulus value information of the color, the spectral reflectance curve obtained by the hyperspectral camera contains detailed information about the object over the entire measured spectral range. This has unique advantages for the color analysis of some special materials or objects. For example, in the field of cultural relics restoration and protection, by analyzing the spectral characteristics of pigments on the surface of cultural relics, we can understand their composition and age information, which provides an important basis for restoration work. In the field of agriculture, the growth status, nutrient content and disease and insect pests of plants can be monitored according to the changes in the spectral reflectance of plant leaves, because the absorption and reflection characteristics of different wavelengths of light will change in different growth stages and health states of plants.
(3) Non-contact measurement
Hyperspectral cameras do not need to make direct contact with the object being measured, which is important in many cases. For some fragile, precious or difficult to reach objects, such as art, cultural relics, biological samples, etc., non-contact measurement can avoid damage or pollution to the object. At the same time, it can also achieve fast, large area color measurement, improve the measurement efficiency. For example, in the color detection of large-scale mural paintings, the color information of the entire mural can be quickly obtained, providing comprehensive data support for protection and restoration work.
四、Experimental test of hyperspectral camera in color measurement
1. Experimental purpose
Test the Lab value of the sample below
2. List of experimental testing instruments
Device name
Model number
Configuration details
Remark
CHNSpec hyperspectral camera
FS-13
Spectral range: 400-1000nm;
Spectral resolution: 2.5nm
Spectral band: 1200
3. Experimental content
The reflectance curve was obtained by external push scan detection of 400-1000nm hyperspectral camera
The experimental measurement process is shown in the figure below:
4. Conclusion
The hyperspectral camera FS-13 was used to shoot the customer's samples, and the Lab value of each sample was obtained from the hyperspectral image analysis, which could be used to replace the color difference meter, and the test stability was good, the sampling position of the test sample was flexible, and multi-point measurement could be made to realize automatic detection.
