Weiping Yang

Multi-spectral imaging using LED illuminations

We present a multi-spectral imaging method, using five types of LED(Light Emitting Diode) light compensated with a tungsten light as the active illuminants, adapting a color camera to sample the spectral image. The imaging model of the multi-spectral imaging system is discussed, and the reflectance is reconstructed using the imaging system matrix derived from the imaging model. The imaging method is simple, low cost and requires no special equipment. The imaging accuracy and stability of the spectral reflectance reconstruction algorithm are discussed through three cases of multi-spectral imaging the Macbeth color checker, a color picture and a leaf infected with powdery mildew disease. It shows that the multi-spectral imaging system can be applied in a variety fields.

Spectrum-based method for quantitatively detecting diseases on cucumber leaf

In plant disease diagnosis, quantitatively detecting the distribution and severity of disease on plant leaf is a fundamental task to design solution and strategy. In this paper, a spectrum based algorithm for leaf disease diagnosis was presented and discussed. This method can locate the disease distribution on leaves and denote the disease signal strength quantitatively by solving linear optimization equation which characterized the categories of the multispectral image pixels. Two typical diseases on cucumber leaves are analyzed and it showed the proposed method can detect the disease area very well. Finally, the properties of the algorithm are also discussed.

Calculation model used for recipe revision in computer color matching

We analyze and discuss the error generated by an algorithm commonly used in computer color matching and point out the necessity of the calculation used for recipe revision. On the basis of this calculation a model for computer color matching used in recipe revision is proposed that will cause the computer color-matching technology to be used more widely and to meet the demands in practice.

Using imaging brightness information to select feature bands for multi-spectral images of cabbage leaves

The experiment takes cabbage leaves as research object to capture images based on multi-spectral imaging system with combination of LCTF and CMOS camera by every 5 nm interval from 400 nm to 720 nm. Firstly, according to the principle of image brightness information, the value of distinguish degree for cabbage leaves are calculated with every band; Then through sorting the value of distinguish degree for cabbage leaves, along with information features of the image and distinguish degree, it can be concluded that band 555 nm, 715 nm, 710 nm, 575 nm, 535 nm, 520 nm, 720 nm, 605 nm and 650 nm have better distinguish degree; Finally, the classification accuracy statistic of feature bands for cabbage leaves is 95.56% and 93.13% through using the principle of Euclidean distance and spectral angle match respectively. It can draw a conclusion that the selected bands are with ideal classification accuracy for cabbage leaves. Therefore, 555 nm, 715 nm, 710 nm, 575 nm, 535 nm, 520nm, 720 nm, 605 nm and 650 nm can be used as feature bands for cabbage leaves. To select feature bands for cabbage leaves is one of effective means to identify the status for plants, which provides a method for fine classification and data processing for plant multi-spectral images with broad prospects and ideal application value.

Sharpness evaluation function of wavelength-related multi-spectral image

In this paper, two objective sharpness functions which results match the subjective result very well are obtained, and it solved the problem of descripting the sharpness property in a multi-spectral image. The multi-spectral image of an international standard vision chart is obtained by a dispersion LCTF multi-spectral imaging system. Subjective experiment is designed and conducted to obtain the sharpness values of the channel image with different light wavelength, and 6 objective sharpness evaluation functions are used to calculate the sharpness/wavelength relationships. Experiment results show that only gray differential function and point sharpness function match the subjective experiment very well, and can be used to describe the sharpness property of the channel images of a dispersion multi-spectral imaging system. Moreover, concerning the computational speed, gray differential function is faster than point sharpness function.

The Spectral Adaptation applied to Image Color Appearance Model

The major aim of Color Appearance Model is trying to resolve the problem of high faithful color reproduction under different viewing conditions, which is one of the main mathematical methods fulfilling the cross-media color production. Currently used Color Appearance Model (CAM) based on von Kries chromatic adaptation transform, that is, CAM acts upon chromatic signals such as tristimulus values rather than spectral characteristics. The spectral adaptation model introduced by Fairchild in 2006 based on the spectral reflectance factor of stimulus and the spectral power distribution function of light source. The new way models the vision phenomenon with no requirement for the chromatic processing. Munsell color chips are used as basic data. Spectral adaptation model and CIELAB color space are employed to predict corresponding color appearance attributes. Then the comparison to CAT02 chromatic adaptation transform is also made. The result shows that spectral adaptation model has a high degree of accuracy in lightness performance, and the relative error is between 2 % and 5 %. But the accuracy of chroma is less than CAT02, and both models have good uniformity in hue performance. Taking the image color appearance model as the carrier, the spectral adaptation model is applied to the image reproduction. The methods are used to model the image with no requirements for the CAT02 chromatic adaptation processing and to use spectral adaptation as pretreatment, then to carry out the chromatic adaptation transformation. Discussion of the advantages and limitations of the model and the practical value in image rendering is also presented.

An infrared imaging computation model and its validation

The infrared imaging equipment is widely used because it can acquire more thermal and material information from infrared band than from visible band. Based on the lighting model which is widely used in computer graphics and the radiation transfer law, a simplified thermal infrared imaging computation model is derived. The following works have been done to derive the model: 1) Adding the surface temperature distribution of the 3D model; 2) Specifying the physical material of the 3D geometry model; 3) Merging the self emitting and the detector response into the imaging model as one term. The ray tracing method is applied to construct an infrared imaging simulation system which can generate the synthetic infrared images of a 3D scene from any angle of view. To validate the infrared imaging computation model, several typical 3D scenes are made, and their infrared images are calculated to compare and contrast with the measured infrared images obtained by a middle infrared band imaging camera. The result shows that: 1) The infrared imaging computation model is capable of producing infrared images which is very similar to those received by thermal infrared camera; 2) The infrared imaging computation model can well simulate the relative brightness contrast in the infrared images, it also can reflect most of the basic infrared imaging characteristics; 3) Some geometry, thermal and material information also can be retrieved from the synthetic infrared images. Quantitative analysis shows that the absolute brightness does not match well, and the reasons are analyzed. By the synthetic infrared images, it also illustrates the difficulty and complexity in infrared image analysis and simulation.

Color difference sensitivity of human vision system for red-green and yellow-blue directions

Purpose: In this paper the chromatic contrast was defined as color difference CIEDE2000, the sensitivity was defined as reciprocal of threshold of the color difference, and the CSFs called color difference sensitivity functions were measured. Methods: The CSFs of 4 subjects were measured for nine spatial frequencies(0.28, 0.56, 1.00, 1.97, 2.95, 4.72, 6.74, 11.80 and 15.74cpd) gratings of mean luminance of 40cd/m2 on a CRT display. Measurements were made for gratings whose average color was a chromatically neutral point(a*=0 and b*=0) and also for modulations around four chromatic points along the color directions a* and b* in the CIELAB color space. Results. The thresholds of color difference are from 0.74 to 6.67 in the range of experimental frequencies. The color difference sensitivity functions are similar with known results that CSF curves for the two chromatic directions are consistently low-pass irrespective of the average color of the stimulus. The sensitivity to gratings for b* direction is identical as for a* direction below spatial frequency 4.72cpd, however, the sensitivity for b* direction is smaller than the one for a* direction above spatial frequency 4.72cpd, which indicates that the CIEDE2000 threshold for grating with lower frequencies(i.e., small color differences) is not related to the chromatic direction and chromatic point of modulation, however, the threshold is related to the chromatic direction and chromatic point of modulation for large color differences.

Fast-scanning spectrophotometry system using a new compensatory model

It is necessary for a modern spectrophotometry system to possess such properties as high measuring accuracy, fast working speed, uniform spectral response in the whole wavelength band, low consumption of power, little heat accruing while working, small size and low weight, etc. To meet all these needs, a spectrophotometry system today usually employs the so-called dual-beam channel configuration in which one sampling channel with another compensatory channel and a cold light source such as a xenon flash lamp, which usually causes an increase of the complexity in the system configurations. To further simplify the instrument configuration, we put forward a new dual-channel compensatory model called the integral compensatory model. Our experimental results show that using this new model, a spectrophotometry system cannot only be greatly simplified in its hardware but can also achieve highly accurate measurements.