Keywords

1 Introduction

With the significant development of military technology in the area of informatization and intelligence, the airborne mission system has increased its information volume in order to adapt to the increasingly complex combat environment. Reasonable color design of the airborne mission system helps to reduce the workload of the operator, improve the efficiency of recognition and cognitive processing, and ultimately serve the operational effectiveness of the system. Closed airborne cabin is an important workplace and operational environment in high-tech warfare, and the harsh airborne environment can aggravate the physiological and psychological load of the worker due to the closed space, loud noise, physical vibration, and complex optical environment conditions.

There have been comprehensive studies on the effects of optical environment on color visual performance [4, 6], therefore this study focused on the issue of vibration and noise. Vibration environment has a significant impact on human visual performance. Wang [5] discovered the vibration frequency and amplitude has a significant impact on the visual task completion time and the difficulty rating, and the direction of vibration has a significant impact on the visual task completion success rate. The study by Mallick [3] points out that in the case of continuous vibrating, choosing the correct color combination of the text and the background can solve the problem of difficulties of data entry task on laptop computers.

Hearing and vision have a clear interaction compensation effect involving the central nervous system. Gai [1] found that noises can significantly affect the subject’s scoring on visual retention, visual response time, curve matching, target tracking, and stereoscopic vision. Liu [2] conducted a visual search task performance study in a speech noise environment.

Most of the researches directly investigated visual ergonomics, and rarely explored the aspect of text color. Moreover, most of them focused on the influence on visual performance of one single environmental variable, and the interaction influence between environmental variables remains unclear. This paper conducted color ergonomics research on harsh environment condition under three task modes, which can help pilot improve information recognition speed and cognitive processing speed, as well as improve their operation efficiency.

2 Experiment 1 - Colored Block Experiment

2.1 Experiment Subject

We have 15 male participants as our subject for the colored-block experiment. Their age is between 21–23, average age \(21.8 \pm 0.6\). They are healthy, and have normal vision or corrected-to-normal vision, normal color vision, and normal color perception.

2.2 Experiment Equipment and Material

Computer. We use Lenovo Erazer Y50-70 Laptop: CPU Intel Core i5 4200H, memory 8 GB DDR3, hard drive 256 GB SSD, Graphics Card Nvidia Geforce GTX 860M/Intel GMA 4600, Screen 15.6\(''\) LCD, resolution \(1920\times 1080\), color depth 32 bit, refresh rate 60 Hz.

Material. Twenty-five colors were chosen from Ostwald Color System (Fig. 1) together with white. Their index number and their RGB values are 0-(230,0,18), 1-(235,97,0), 2-(243,152,0), 3-(252,200,0), 4-(255,251,0), 5-(207,219,0), 6-(143,195,31), 7-(34,172,56), 8-(0,153,68), 9-(0,155,107), 10-(0,158,150), 11-(0,160,193), 12-(0,160,233), 13-(0,134,209), 14-(0,104,183), 15-(0,71,157), 16-(29,32,136), 17-(96,25,134), 18-(146,7,131), 19-(190,0,129), 20-(228,0,127), 21-(229,0,106), 22-(229,0,79), 23-(230,0,51), 24-(255,255,255). Colored block’s size is 2.5 cm \(\times \) 2.5 cm. Experiment interface’s background color is black, its RGB value (0,0,0).

Fig. 1.
figure 1

Ostwald Color System and According Parameters (Color figure online)

Full size image

2.3 Experiment Design

Experiment program is coded with C++ and Qt framework version 5.11. Before experiment starts, the subjects are sufficiently practiced in order to avoid the practice effect. During the experiments, the indoor light environment stays unchanged, subjects are seated on the chair 60 cm away from the display, both eyes keep parallel with the screen. The subject uses standard computer hardware such as standard mouse and standard keyboard to interact with computer.

Fixed-size colored block was randomly displayed in 25 colors at the center of the screen. Each subject was asked to look at the center of the screen, and press the space bar as soon as they noticed and recognized the color. The colored block will not disappear until the subject has reacted. The interval of the appearance of each colored block was random, ranging between 1–3 s. Each color was repeated 4 times, and each subject performed a total of 100 reactions. Their reaction time were recorded. During the experiment, the subject must be focused, and a reaction is only recorded if it is made after the colored block appeared, meaning pressing the space bar in advance is prohibited and also useless.

2.4 Experiment Results

We analyze the results from the experiment of the reaction time. The abnormal values are omitted from the analysis, then the mean value of each color are calculated and shown in Fig. 2.

Fig. 2.
figure 2

Mean reaction time of each color in experiment 1 (Color figure online)

Full size image

From Table 1, analysis of variance showed that there was a significant difference in reaction time of the 25 colors \((F=1.84, P<0.05)\). Least Significant Difference (LSD) of the 25 colors showed that, the blue-violet color series (No. 15, 16, 17) have performance significantly slower than the yellow color series. Also, red, yellow, cyan and white has the tendency of faster than other colors. The difference in other colors’ performance is not significant.

Table 1. ANOVA: the effect of colors on completion time
Full size table

Based on these results, also considering the integrity of color series, five colors were chosen as the color for the interface in the following experiment 2: red (230,0,18), yellow (252,251,0), green (0,155,107), blue (0,134,209) and white (255,255,255).

We also studied the count of the color with the shortest reaction time (Fig. 3) and longest reaction time (Fig. 4) within each experiment group (of 25 experiments). In the shortest reaction time count, the top color is yellow (5 times), accounting for 12% of the total, followed by white. In the longest reaction time count, blue color series (dark blue, light blue, sky blue) are the most (12 times), accounting for 30% of the total.

Fig. 3.
figure 3

Count of shortest reaction time for each color (Color figure online)

Full size image
Fig. 4.
figure 4

Count of longest reaction time for each color (Color figure online)

Full size image

3 Experiment 2 - Color Ergonomics Experiment

3.1 Experiment Subject

We have 10 male participants as our subject for this experiment. Their age is between 21–23, average age \(22.3 \pm 0.4\). They are healthy, and have normal vision or corrected-to-normal vision, normal color vision, and normal color perception.

3.2 Experiment Equipment and Material

Computer. We use Lenovo Erazer Y50-70 Laptop: CPU Intel Core i5 4200H, memory 8 GB DDR3, hard drive 256 GB Solid State Drive, Graphics Card Nvidia Geforece GTX 860M/Intel GMA 4600, Screen 15.6\(''\) LCD, resolution \(1920\,\times \,1080\), color depth 32 bit, refresh rate 60 Hz.

Material. Experiment material used in this experiment are shown in Fig. 5. The three parts on the interface are (from left to right respectively) search task, calculation task and monitor task. We use five colors chosen from Experiment 1 as the interface color: red (230,0,18), yellow (252,251,0), green (0,155,107), blue (0,134,209) and white (255,255,255). The noise is simulated by the software Soundmasker.

3.3 Experiment Design

Experiment program is coded with C++ and Qt framework version 5.11. Before experiment starts, the subjects are sufficiently practiced in order to avoid the practice effect. During the experiments, the indoor light environment stays unchanged, subjects are seated on the chair 60 cm away from the display, both eyes keep parallel with the screen. The subject uses standard computer hardware such as standard mouse and standard keyboard to interact with computer.

During the experiment, the subject needs to complete three type of tasks.

  • Search task (T1): 25 randomly-chosen different English letters will appear in a \(5\times 5\) matrix in random order, and one of the letters will be randomly chosen as the target letter. Subject should find and click the target letter in the matrix as soon as possible.

  • Calculation task (T2): Two-digit to two-digit add and subtract tasks will be randomly generated and displayed on the interface. The subject need to input the calculated result by keyboard and click “enter” to submit the answer.

  • Monitor task (T3): The program will randomly display words from four major categories (animals, plants, food, city names) in a \(5\times 3\) matrix. We randomly chose one category as the target category. During each task, there will be only zero or one word of the target category. Subjects need to monitor if any word from the target category is shown in the matrix. If there is, the subject needs to click on it; if not, the subject do not need to react.

To simulate different environment scenarios, the experiment designed different vibration levels and noise levels. Vibration level includes no vibration and high vibration (frequency 20 Hz, amplitude 2 mm) to the graphic interface. Noise level includes low noise (45 dB white noise) and high noise (80 Hz white noise).

Each subject was asked to complete tasks in two environments.

  • General Environment (E1): low noise (45 dB) with no vibration.

  • Harsh Environment (E2): loud noise (80 dB) with high vibration (frequency 20 Hz, amplitude 2 mm).

Fig. 5.
figure 5

Graphic interface of experiment 2 (Color figure online)

Full size image

There were five task groups in each environment corresponding to the five colors selected from Experiment 1. There were four task page within each task group, and each task page consisted of a search task, a calculation task, and a monitor task. The operation completion time and success rate of the subject were recorded.

Therefore overall, each subject needs to complete 40 search tasks, 40 calculation tasks and 40 monitor tasks. For monitor tasks, however, the subject need not to react if there are no words from the target category, therefore no reaction time on these tasks. We programmed 10 out of 40 tasks to contain words from the target category. That means, for each subject we have the experiment results for 40 search tasks, 40 calculation tasks and 10 monitor tasks.

3.4 Experiment Results

The mean completion time of each task under different environments is shown in Fig. 6. The analysis of variance is shown in Table 2.

Fig. 6.
figure 6

Mean value of completion time of experiment 2 (Color figure online)

Full size image

General Environment (Low Noise, No Vibration). In search task, there was significant difference in performance for different colors \((F = 16.122, P < 0.05)\). Combined with Least Significant Difference (LSD), the results showed that yellow was significantly better than other colors. In calculation task, there was a trend of significant difference \((F = 1.999, P < 0.1)\). In monitor task, there was no significant difference in completion time between colors.

Harsh Environment (Loud Noise, High Vibration). In search task, there was significant difference in performance for different colors \((F = 29.163, P < 0.05)\). Combined with histogram and Least Significant Difference (LSD), the completion time of yellow and white was significantly shorter than that of blue and green, and there was no significant difference between red and white. The completion time ranking from fastest to slowest is yellow, white, red, blue, and green. In calculation task, there was a significant difference in performance for different colors \((F\,=\,2.511, P < 0.05)\). Blue and green were significantly slower than the other three colors, while the other colors were not significantly different. In monitor task, there was no significant difference in completion time between colors.

Table 2. ANOVA: the effect of colors on completion time, grouped by environment and task
Full size table

Combined with Table 3 and LSD results, the differences between colors in the general environment and the harsh environment are both significant, and the difference is slightly larger when in the harsh environment. The multivariate analysis of variance (Table 5) also showed that the interaction of the color and environment had a significant effect on the completion time \((F=3.345, P<0.05)\).

Table 3. ANOVA: The effect of colors on completion time, grouped by environment
Full size table

Combined with Table 4 and the LSD results, the difference between colors was significant in Task 1 \((F=38.244, P<0.05)\), while the difference between Task 2 and Task 3 was not significant. The multivariate analysis of variance (Table 5) also showed that the interaction of color and task type had a significant effect on the completion time \((F=2.866, P<0.05)\).

Table 4. ANOVA: The effect of colors on completion time, grouped by task
Full size table
Table 5. The multi-factor variance analysis of the influence of color, environment and task on completion time
Full size table

4 Conclusion

Color Performance. For different tasks and different environments, color performance has a clear trend. As for the speed of recognition process and cognitive process, yellow and white performs considerably better than blue and green, and red is of medium performance. This indicates that. Both experiment 1 and 2 showed that the blue has the lowest performance in recognition and cognitive.

Impact of Color in Different Environments. The selection of color has less effect in the general environment than in harsh environment. When the environment worsened, people are more sensitive to color, and the differences between the performances of colors are magnified. This shows that in severe environments, the choice and application of color needs to be made with more caution.

Impact of Color on Different Task Modes. Color selection has a significant impact on the search task, but not on calculation and monitor tasks. A search task is a process of identification and matching which involves little processing, while the calculation and monitor tasks have longer information processing time, thus the effect of color is weakened. Therefore, when applying of principles of color selection, it is best to focus on recognition rather than cognition. Color selection should be designed with great effort for recognition tasks.