Open Access
Add to BibSonomy
Abstract
A method was developed to investigate the reflected properties of LED and HPS-lighted wet roads using on-site measurement of luminance images. An ILMD was used to obtain high observation angles by placing it at a close distance to the ROI. The luminance images were analyzed to derive a distribution of reflected luminance as a function of deviation, incidence, and observation angles. Furthermore, an investigation of the reflected characteristics of a dry asphalt road illuminated by LED lighting was conducted, employing on-site measurements of the images of luminance and the distribution of illuminance. The results provide valuable insights for designing lighting systems and mitigating the adverse effects of glare and specular reflection on road safety.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Reflectance is an important optical characteristic of roads, as it determines the amount of light that is reflected back to the driver's eye. The higher the reflectance of the road surface, the more visible it is to the driver, especially in low-light conditions [1]. Reflectance also plays a critical role in the visibility of road markings, such as lane lines, arrows, and symbols [2–4]. Luminance (L) is an important optical characteristic of roads, as it determines how bright a road surface appears to the driver. The luminance of a road surface can be influenced by two main factors: the level of ambient light present in the surrounding environment and the reflective properties of the road surface itself. [5,6]. In general, road surfaces with higher luminance values are more visible to the driver, especially in low-light conditions. The illumination (E) of roadways is an important aspect of road safety, as it enables drivers to see and navigate the road surface [7–9]. Luminance can be influenced by factors such as the incidence angle the illuminance ($\varepsilon $), the angle of observation ($\alpha $), and deviation angle ($\beta $). To simulate the characteristic of a road lighting environment, engineers often use standardized reduced luminance coefficients, or r-tables, which provide average luminance coefficients for different classes of road surfaces under dry and wet conditions [10]. The r-tables provide valuable information on the reflectance properties of road surfaces under different lighting conditions, which can be used in roadway lighting simulations to optimize the placement and luminance intensity of lighting fixtures for improved visibility and safety [11,12]. Nevertheless, the applicability of r-tables is restricted by specific limitations, including restrictions on low-observation angles ($0.5^\circ < \alpha < 2^\circ $), straight-road conditions, and knowledge of the road material and physical state. Moreover, the reflection properties of a real-world road may not always provide optimal visual conditions for the driver [13].
Taiwan and other areas of Asia are situated in a climate characterized by hot and humid summers (up to 33 ℃ and 80%RH), typhoon attacks, and cool and damp winters (up to 8 ℃ and 80%RH) [14]. This leads to numerous weather challenges for LED road lighting, particularly for highways where the safety and comfort of road users are essential [15]. As road users travel along the road, one crucial indicator of driving safety under dry, wet, and rainy road conditions is the reflectivity of the road surface. When rainwater falls on a road surface, it forms a thin water film that can reduce the visibility of the road surface [16,17]. Additionally, there is a frequent comment that the reflection from wet LED-lighted roads is more glaring than that from roads lighted by traditional luminaires, mainly when observed close to the lighting pole due to the high observation angle [18]. Also, rainwater on the road surface can reduce the visibility of road markings for drivers. Under wet conditions, a water film on the road can diminish the visibility and differentiation of the road markings from the road surface [19]. However, some road markings are designed to have higher reflectance even in rainy conditions. For example, some road markings are made with retro-reflective materials that reflect light back to the driver's eyes, increasing visibility even in rainy conditions. Recent advancements in NIR camera technology have also enabled moisture detection on road surfaces. This is particularly useful for monitoring road conditions in rainy or snowy weather, where the presence of moisture can affect the safety and performance of the road surface [20].
Road lighting is essential for ensuring road safety in rainy conditions, but it can also cause glare, reducing visibility and increasing the risk of accidents. The generation of glare from road lighting can be influenced by various factors, such as the luminaire's design, the placement of the lighting, and the type of light source utilized. Strategies such as appropriate lighting design, luminaire placement, and optimizing the luminance intensity of road lighting and correlative color temperatures can help mitigate the effects of glare and improve visibility in rainy conditions. A quantitative study and comparison of reflections from wet roads lighted by LED and traditional luminaires are needed. Understanding these optical characteristics can help road designers and engineers create safer roads and improve transportation safety in rainy conditions. Therefore, the establishment of measurement and analysis techniques for road lighting luminance and pavement characteristics can be used to assess the quantitative reflective properties of road lighting and its correlation to the experience of road users under LED lighting.
For vehicle-to-vehicle (V2V) visible light communication (VLC) systems, it is crucial to consider the reflectance of the road surface when designing and deploying these systems. Road surfaces that have higher reflectivity can be utilized in areas with high-traffic-density to enhance the effectiveness of vehicle-to-vehicle (V2V) visible light communication (VLC) systems, especially the communication link's signal-to-noise ratio (SNR) [21].
We first established a weather testing system and technology to create dry and wet road surface environments, causing the road surface to produce diffuse, specular, or near-specular reflection in response to road lighting. The reflection characteristics are related to the direction of the light source along the road surface, the size of the light source, and the type of light source. Then, the uniformity contrast characteristics of the road surface illumination and luminance distribution, as well as the reflection or scattering of the road surface, are measured and analyzed. This study compared two types of illumination, LED and HPS road lighting, specifically in wet road lighting environments. By evaluating their performance in terms of road properties and lighting effectiveness, we aimed to gain a comprehensive understanding of their color properties and glaring effects. Additionally, as part of our research, we proposed on-site measurement methods for assessing road properties in both wet and dry asphalt conditions. These methods contribute to the development of standardized approaches for evaluating road surfaces and their impact on road safety and visibility.
2. Measurement method of ILMD for road reflection characteristics
The following experiment content and result analysis describe the high-angle reflection of LED road lighting on extremely wet road surfaces and high-angle reflection on dry road surfaces. The luminance images and illuminance distribution were subjected to analysis and converted into a luminance coefficient represented by the equation ($q = L/E$), where q is a function of β, ε, and α. These findings and procedures provide more physical properties that can be used for designing road light environments that prioritize safety.
A diagram of the experimental two-lane road in southern Taiwan is presented in Fig. 1. The lighting system on the road consisted of LED luminaires mounted on lighting poles (indicated by diamond symbols in Fig. 1) along the road. The LED luminaires and high-pressure sodium (HPS) luminaires were installed with the same configuration, including a height of 10 m, lane width of 3.8 m, and a distance of 43.8 m between them, to allow for accurate measurements and comparisons.
Fig. 1. The diagram of (a) the experimental road, (b) the angular relationships for observer and road lighting, and (c) the sampling points of luminance distribution.
The road's luminance was evaluated using a calibrated imaging luminance measuring device (ILMD) equipped with a (10 to 22) mm focal length lens, used for low luminance (< 1 cd/m2), far distances (> 60 m), and small acceptance angles (< 0.1°) [22]. The related calibration and usage techniques of ILMD can be referred to in the literature. The ILMD was placed at distances (D) from 2 m to 12 m, at the 2 m interval, green cross symbols in Fig. 1, from the nearest light pole, and the ILMD set at a height of 1.5 m. The ILMD has a requirement for an acceptance angle of less than 0.1°. The analysis data presented in this study is derived from measurements obtained within small acceptance angles. Specifically, the data was integrated from the regions of interest (ROIs) where the acceptance angles ranged between 0.3° to 0.8°. The sampling points analyzed were marked with solid circles in Fig. 1(a). The angular relationships and the position of calculation points for luminance distribution were shown in Fig. (b) and Fig. (c), respectively. The illuminances on the sampling points were automatically measured by an array-type illuminance measurement system composed of distributed illuminance meters and a distance counter [23,24].
3. Reflection properties of a wet road lighting environment
Firstly, we can discuss the situation where the road surface was wet but free of standing water. A glossy reflection will be mixed with a diffuse reflection on the road surface. As drivers feel that this kind of road surface reflection interferes with their vision, it mainly occurs in high-angle situations at close distances. Also, the image luminance on the road surface is measured [25].
Figure 2 shows the luminance reflection image of LED road lighting on a wet road surface, which is obtained using a lens with a 10 mm ultra-wide-angle. The small circular area corresponds to the solid circle position in Fig. 1, which is used to analyze the luminance reflection at these positions. The coordinate origin is near the base of the light pole. These images show that the specular reflection is distributed in a slender shape along the line of sight, and its peak position and distribution change with the relative position of the observer, road lighting, and optical properties of the road. The luminance distribution of a road surface is interrelated with various factors, such as the uniformity and wetness of the road surface, the type and usage of the road surface material, and the distribution of illuminance provided by the road lighting.
Fig. 2. The luminance images of LED luminaires captured by ILMD at (a) D = 2 m, (b) D = 6 m, (c) D = 10 m.
The luminance distribution of the longitudinal and transverse sampling points near the center of the road are depicted in Fig. 3(a) and Fig. 3(b), respectively. The position of the observation point D affects the peak position of the specular reflection, and as D changes, the peak position also changes. A small local peak can be observed around x = 8 m in Fig. 3(a), the sampling points indicated by the “X” symbol in Fig. 1, but it does not significantly impact the subsequent analysis. On the other hand, Fig. 3(b) reveals that the peak position of the specular reflection changes due to the curved uphill section of the road as D changes, the sampling points indicated by the “+” symbol in Fig. 1. The maximum and minimum luminance values are 98.6 cd/m2 and 0.18 cd/m2, respectively, with a ratio of 560:1. These observations demonstrate the high variability of the luminance distribution along the wet road, and suggest the importance of considering the position of the observation point when studying the specular reflection of light.
In this study, we aimed to obtain more physical properties related to road surface reflection. To achieve this, the data previously collected were converted to luminance as a function of deviation angle, incidence angle, and observation angle ($\beta $, $\varepsilon $, and $\alpha $), in accordance with the definition provided by CIE-140:2019. The observation angles obtained ranged from 1.5° to 33.8°, which covers the typical field of view of a passenger vehicle driver and is not limited to the assumption of $\alpha < 5^\circ $ for r-tables. The obtained data were then analyzed by plotting the normalized luminance ($L/{L_{\textrm{max}}}$) and Color Correlative Temperature (CCT) as functions of $\beta $ and $\alpha + \varepsilon $. The results revealed that the distribution of the data is near an oblong bell shape with a peak at around $\beta = 0^\circ $ and $\alpha + \varepsilon = 87^\circ $, which deviates slightly from pure specular reflection of $\alpha + \varepsilon = 90^\circ $, shown in Fig. 4 and Fig. 5.
Fig. 4. Normalized luminance as functions of $\beta $ and $\alpha + \varepsilon $ for LED lighting; (a) functions of $\beta $, (b) function of $\alpha + \varepsilon $ .
Fig. 5. CCT as functions of $\beta $ and $\alpha + \varepsilon $ for LED lighting; (a) functions of $\beta $, (b) function of $\alpha + \varepsilon $ .
The same experiment was conducted by replacing HPS lighting on the original lamppost to compare with LED lighting. Figure 6 shows an example of the reflection luminance image of HPS road lighting on a wet road surface. From these figures, it can be seen that the specular reflection is also distributed in a slender shape along the line of sight, and the peak value of the specular reflection changes more slowly than that of LED lighting. Another local peak value appears at around x = 12 m, but it is not obvious. The peak value of specular reflection also changes with the change of the observation position D. The maximum and minimum values of luminance are 42.6 cd/m2 and 0.14 cd/m2, respectively, with a ratio of 295:1, which is lower than that of LED. There is a significant difference between the two when observed on site, but which numerical value will cause discomfort needs further studies. By converting the coordinates of the location points of all sampling points of the image luminance under each observation distance D into angular coordinates, the logarithmic normalized luminance $L/{L_{\textrm{max}}}$ and the relationship between the angle coordinate $\beta $ and $\alpha + \varepsilon $ can be obtained and summarized in Fig. 7(a) and Fig. 7(b). As shown in Fig. 8 is the CCT as the function of $\beta $ and $\alpha + \varepsilon $.
Fig. 6. The luminance images of HPS luminaires captured by ILMD at (a) D = 2 m, (b) D = 6 m, (c) D = 10 m.
Fig. 7. Normalized luminance as functions of $\beta $ and $\alpha + \varepsilon $ for HPS lighting; (a) functions of $\beta $, (b) function of $\alpha + \varepsilon $.
Fig. 8. CCT as functions of $\beta $ and $\alpha + \varepsilon $ for HPS lighting; (a) functions of $\beta $, (b) function of $\alpha + \varepsilon $.
In this study, the data of $L/{L_{\textrm{max}}} > 10\; \%$ in Fig. 4 and Fig. 7 were analyzed and plotted in Fig. 9 to investigate the impact of LED and HPS lighting on reflected luminance in wet road conditions. The results show that the region corresponding to LED lighting is an oblong oval bounded between $- 3^\circ < \beta < 2^\circ $ and $78^\circ < \alpha + \varepsilon < 92^\circ $. In contrast, the region for HPS lighting is between $- 2^\circ < \beta < 6^\circ $ and $64^\circ < \alpha + \varepsilon < 95^\circ $, which is broader than for LED lighting. These findings suggest that, in wet road conditions with a high observation angle ranging between 1.5° and 35°, the reflected luminance by LED lighting is more concentrated and has higher contrast than those by HPS. This provides quantitative evidence for the observation that the reflected light from a wet road is more specular and has more glare under LED lighting than traditional luminaires. Overall, this study provides important insights into the impact of LED and HPS lighting on reflected luminance in wet road conditions, which could have implications for road safety and lighting design.
Fig. 9. Distribution of $L/{L_{\textrm{max}}} > 10\; \%$. Solid circles and empty circles are corresponding to LED and HPS lightings, respectively.
4. Dry asphalt road measurement result
Usually, the road lighting luminance on dry road surfaces is carried out using r-tables with low viewing angles ($0.5^\circ < \alpha < 2^\circ $) [26], and high angles are often neglected. However, at such viewing angles, the driver is already more than 60 m away from the observed road surface (calculated based on a driver's height and the tangent angle), which does not match the actual driving experience very well. Therefore, this study makes use of the convenience of ILMD to perform close-range high-angle luminance measurements on actual roads and converts the spatial coordinates to angular coordinates to investigate changes in luminance and illumination ratio at high angles.
In the experiment, only one LED lighting was turned on, and this configuration is similar to Fig. 1 in CIE-140:2019. The ILMD was placed 1.5 m above the ground at the center of the outer lane, and its longitudinal distance D from the lamp post, which was the gray bar shown in Fig. 10. Figure 10 shows typical luminance images and sampling points for D = 2 m, 4 m, 8 m, and 10 m. The region of interest, indicated by green circles, were chosen for both luminance and illuminance measurements based on the guidelines outlined in CIE-140:2019. The notation W and S correspond to the transverse and longitudinal direction of an ROI, respectively.
Fig. 10. The luminance distribution captured by ILMD at D: (a) 2 m, (b) 4 m, (c) 8 m, (d) 10 m, and (e) the green circles are the ROIs.
Figure 11(a) shows the horizontal illuminance distribution along the road in Fig. 10, where the solid circle, empty circle, and solid triangle represent the positions with a width of W = 0.62 m, 1.88 m, and 3.13 m along the road, respectively. Figure 10(e) shows the central pavement reflectance distribution along the road in Fig. 11, where the solid circle, empty circle, solid triangle, empty triangle, and solid square represent the positions at observation distances of D = 2 m, 4 m, 6 m, 8 m, and 10 m, respectively.
Fig. 11. The illuminance and luminance distribution; (a) the illuminance was plotted as a function of the longitudinal distance S, (b) the luminance distribution measured in various observation distances D at W = 1.88 m.
The obtained data were utilized to calculate the luminance coefficient q and derive the angles $\beta $, $\varepsilon $, and $\alpha $ based on the transverse distance (W), longitudinal distance (S), and observer's distance from the nearest light pole (D), using established angular relationships. The measured $\varepsilon $ angles varied from 2° to 27°, and the measured $\alpha $ angles ranged from 10° to 43°. Furthermore, the angle $\varphi $, which represents the angle between the incident and observation directions, was computed using the following equation
The relationship between luminance with $\varphi $ and $\beta $ is shown in Fig. 12, where the solid circle, empty circle, and solid triangle represent the positions along the width direction W = 0.62 m, 1.88 m, and 3.13 m, respectively. It can be seen that the closer to the center of the road, the luminance is higher, but the feature is still not very regular. Dividing the luminance data by the horizontal illuminance, the luminance coefficient $q = L/E$ was obtained and shown in Fig. 13 as functions of $\varphi $ and $\beta $. Note that the observation angle ranges from 2.5° to 26.9°, and is much wider than in usual experiments. It can be seen that the luminance ratio as functions of $\varphi $ and $\beta $ approximately exhibit a parabolic relationship, so regression analysis is carried out using the following equation:
The fitted parameters are listed in Table 1. The Fig. 13(b) is one of the projection images of the function of (r, $\varphi $, $\beta $). The ${\varphi _\textrm{0}}$ and ${\beta _0}$ are the parameters of curve fitting by Eq. (2). It can be seen that the minimum value of the luminance coefficient of this road surface is between 0.04 and 0.05, and the angle ${\varphi _\textrm{0}}$ is approximately between 60° and 74°, while the angle ${\beta _0}$ is approximately between 90° and 214°, depending on the lateral position W. The differences in these parameters may be due to the long-term wear and tear of the road surface at different positions. These results comprehensively analyze the old asphalt road surface, with curved uphill roads observed at large angles. It is expected that using the new measurement and analysis method presented in this section, the reflective characteristics of road surfaces under LED lighting can be obtained in a practical and novel way for other materials.
Table 1. The parameters of q, $\varphi $, and $\beta $.
5. Conclusion
This article presents a novel approach to measuring and analyzing the reflective properties of LED road lighting on the road surface. Based on the ILMD and illumination distribution automatic measurement method for road lighting on site, the spatial distribution of road surface reflection luminance is studied. The technique of on-site measurement and analysis presented in this article has the potential to be utilized on different roads featuring diverse pavements, physical attributes, types of road lights, and observation angles. Its application can result in the acquisition of a practical luminance coefficient that can be used to enhance the lighting design of existing roads.
The observation angles ranging from 1.5° to 33.8°, as measured by ILMD, can be utilized to evaluate the wet road surface under various lighting conditions. This study focuses on analyzing the glaring effect through the examination of luminance distribution at high observation angles.
Based on the CIE-144:2001 classification, the average luminance coefficient of the experimental road surface falls within the range of C2 and R2. Specifically, the average luminance coefficient of 0.07 was obtained at low viewing angles (0.5° < α < 2°). The measured luminance coefficient of the experimental road surface using ILMD also aligns closely with the values outlined in CIE-144:2001, with a recorded value of 0.07.
Using this technique in the LED lighting experimental fields can obtain physical parameters that are difficult to measure and analyze with conventional methods, such as the luminance contrast of wet or dry road surfaces, especially the LED lighting road surface reflection glare caused by high angle incidence and reflection that is of great concern.
Funding
Ministry of Economic Affairs (112-1403-10-23-01).
Acknowledgments
The author would like to thank Breault Research Organization (BRO), Inc. for sponsoring ASAP software program.
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. C. D. Galatanu and L. Canale, “Measurement of Reflectance properties of asphalt using photographical methods,” Proc. 2020 IEEE International Conference on Environment and Electrical Engineering and 2020 IEEE Industrial and Commercial Power Systems Europe (EEEIC / I&CPS Europe), 1–6 (2020).
2. R. M. Spieringhs, J. Audenaert, K. Smet, I. Heynderickx, and P. Hanselaer, “Road marking BRDF model applicable for a wide range of incident illumination conditions,” J. Opt. Soc. Am. A 40(3), 590–601 (2023). [CrossRef]
3. S. Y. Shin, J. I. Lee, W. J. Chung, S.-H. Cho, and Y. G. Choi, “Assessing the refractive index of glass beads for use in road-marking applications via retroreflectance measurement,” Curr. Opt. Photonics 3(5), 415–422 (2019). [CrossRef]
4. T. Schnell, F. Aktan, and Y. C. Lee, “Nighttime visibility and retroreflectance of pavement markings in dry wet and rainy conditions,” Transp. Res. Rec. 1824(1), 144–155 (2003). [CrossRef]
5. J. Hu, S. Sun, and R. Wang, “Research on the influence of light source characteristics on traffic visual distance in foggy areas at night,” Build Environ. 212, 108818 (2022). [CrossRef]
6. C. C. Sun, X. H. Lee, I. Moreno, C. H. Lee, Y. W. Yu, T. H. Yang, and T. Y. Chung, “Design of LED street lighting adapted for free-form roads,” IEEE Photonics J. 9(1), 1–13 (2017). [CrossRef]
7. T.-C. Teng, W.-S. Sun, and J.-L. Lin, “Designing an LED luminaire with balance between uniformity of luminance and illuminance for non-Lambertian road surfaces,” Appl. Opt. 56(10), 2604–2613 (2017). [CrossRef]
8. A. Zatari, G. Dodds, K. McMenemy, and R. Robinson, “Glare, luminance, and illuminance measurements of road lighting using vehicle mounted CCD cameras,” LEUKOS 1(2), 85–106 (2005). [CrossRef]
9. Z. Zhu, D. Ma, Q. Hu, Y. Tang, and R. Liang, “Catadioptric freeform optical system design for LED off-axis road illumination applications,” Opt. Express 26(2), A54–A65 (2018). [CrossRef]
10. “Road Lighting Calculations,” CIE 140:2019.
11. V. Boucher and F. Greffier, “Space of basis functions to retrieve road surface reflection tables from luminance maps,” Light. Res. Technol.2022;0(0). [CrossRef]
12. V. Muzet, J. Bernasconi, P. Iacomussi, S. Liandrat, F. Greffier, P. Blattner, J. Reber, and M. Lindgren, “Review of road surface photometry methods and devices–Proposal for new measurement geometries,” Light. Res. Technol 53(3), 213–229 (2021). [CrossRef]
13. A.-M. Ylinen, T. Pellinen, J. Valtonen, M. Puolakka, and L. Halonen, “Investigation of Pavement light reflection characteristics,” Road Mater. Pavement Des. 12(3), 587–614 (2011). [CrossRef]
14. L. W. Lai, “Poor visibility in winter due to synergistic effect related to fine particulate matter and relative humidity in the Taipei metropolis, Taiwan,” Atmosphere 13(2), 270 (2022). [CrossRef]
15. A. K. Jägerbrand and J. Sjöbergh, “Effects of weather conditions, light conditions, and road lighting on vehicle speed,” SpringerPlus 5(1), 505 (2016). [CrossRef]
16. N. Hautière, E. Dumont, R. Brémond, and V. Ledoux, “Review of the mechanisms of visibility reduction by rain and wet road,” In 8th International Symposium on Automotive Lighting, 445–455 (2009).
17. R. Kurata, H. Watanabe, M. Tohno, T. Ishii, and H. Oouchi, “Evaluation of the detection characteristics of road sensors under poor-visibility conditions,” In IEEE Intelligent Vehicles Symposium, 538–543 (2004).
18. M. S. Shehata, J. Cai, W. M. Badawy, T. W. Burr, M. S. Pervez, E. J. Johannesson, and A. Radmanesh, “Video-based automatic incident detection for smart roads: The outdoor environmental challenges regarding false alarms,” IEEE Trans. Intell. Transport. Syst. 9(2), 349–360 (2008). [CrossRef]
19. T. Yager, B. Gallaway, D. L. Ivey, and J. M. Mounce, “Influence of Roadway discontinuities on safety: Water accumulation,” (Transportation Research Board of the National Academies, 2009), pp. 51–60.
20. P. Jonsson, J. Casselgren, and B. Thörnberg, “Road surface status classification using spectral analysis of NIR camera images,” IEEE Sens. J. 15(3), 1641–1656 (2015). [CrossRef]
21. F. M. Alsalami, O. C. L. Haas, A. Al-Kinani, C.-X. Wang, Z. Ahmad, and S. Rajbhandari, “Impact of dynamic traffic on vehicle-to-vehicle visible light communication systems,” IEEE Syst J. 16(3), 3512–3521 (2022). [CrossRef]
22. “On site measurement of the photometric properties of road and tunnel lighting,” CIE 194:2011.
23. “Road Lighting - Part 4: Methods of Measuring Lighting Performance,” CEN EN 13201-4:2003.
24. S. W. Hsu, K. N. Wu, and S. T. Hung, “Performance of led road lightings studied by detailed in-field measurements with various devices”, Proceedings of CIE 2015 (2015).
25. S. W. Hsu, C. H. Chen, and Y. D Jiaan, “Measurements of UGR of LED light by a DSLR colorimeter”, in SPIE Optical Engineering + Applications (2012), pp. 848415.
26. “Road Surface and Road Marking Reflection Characteristics,” CIE 144:2001.
