Colorimetric Compensation in Video Mapping for Luggage Inspection

elan Poyer

1,2

and Christophe Hurter

Ecole Nationale de l’Aviation Civile, Toulouse, France

Thales AVS, V

elizy-Villacoublay, France

Keywords:

Video Mapping, Colorimetric Compensation, Luggage Inspection, Interpolation.

Abstract:

Airport luggage inspection agents work under time pressure to identify and localize dangerous items using

three-dimensional scans. Video mapping can signiﬁcantly enhance speed and efﬁciency by projecting real-

time information helping to localize threats to be removed. However, maintaining color ﬁdelity is crucial, as

accurate color representation provides key information for decision-making. Previous research has explored

color correction techniques for complex surfaces, but these often require extensive calibration, limiting their

real-time applicability. Our approach addresses this limitation by using a pre-recorded database to maintain

color compensation without the need for frequent recalibration. We built this colorimetric database that records

how surfaces with similar textures reﬂect colors. Using Shepard’s interpolation, our algorithm generalizes the

color correction to new surfaces with similar textures, allowing for real-time adjustments without interrupting

workﬂow. This paper aims to lay the foundation for large-scale studies. The results show good performance for

hues such as orange but the method’s effectiveness varies across the color spectrum, with limited improvements

on blue hues due to predictable losses in luminance and saturation. This highlights the need for new techniques

to overcome the physical limitations of projectors.

1 INTRODUCTION

Luggage inspection is a safety-critical environment,

meaning that operations are carried out under signif-

icant pressure to achieve speciﬁc tasks within strict

time limits. These settings often involve high-stakes

decision-making. Delays or inefﬁciencies in such

contexts can have serious consequences, including

threats to security, safety, or operational efﬁciency.

Security checkpoint operators use three-dimensional

scans of luggage to identify and localize threats or

dangerous items. This highlights the need to clearly

differentiate the virtual representation of luggage, de-

rived from scans, from the physical luggage being

inspected. The transfer function renders the vir-

tual volume of the luggage based on its X-ray ab-

sorbance (Drebin et al., 1988; Metz and Doi, 1979;

Kindlmann, 2002). Therefore, colors provide key in-

formation about the materials constituting the objects,

aiding in their recognition—Orange for organics,

green for inorganics and blue for metals (Figure 1).

Under these high-pressure conditions, the preattentive

processing (Treisman and Gelade, 1980) allows oper-

ators to instantly differentiate metal, which is often

the material constituting dangerous items, from other

materials. This ties into Bertin’s principles of semi-

ology of graphics (Bertin, 1983), where visual vari-

ables like color are essential for effective communi-

cation and interpretation of information in complex

visual environments. Furthermore, this also corre-

sponds to Gibson’s concept of ecological design (Gib-

son and Pick, 2000), which emphasizes that visual

cues should be clearly perceptible and aligned with

the task at hand to facilitate intuitive decision-making.

If the displayed colors are inaccurate, it can lead to

misinterpretation and slow the decision-making pro-

cess. Thus, maintaining precise color accuracy in crit-

ical applications is essential for ensuring effective and

reliable outcomes in these high-stakes environments.

Figure 1: Baggage Scanner Output with Color Mapping.

Colors represent different materials based on their density

and composition.

Poyer, M. and Hurter, C.

Colorimetric Compensation in Video Mapping for Luggage Inspection.

DOI: 10.5220/0013312200003912

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 20th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2025) - Volume 1: GRAPP, HUCAPP

and IVAPP, pages 947-954

ISBN: 978-989-758-728-3; ISSN: 2184-4321

947

In such environments, real-time processing, rapid

adaptability, and high accuracy are essential, as even

minor errors or small loss of concentration can have

cascading effects, potentially compromising the mis-

sion’s objectives or endangering lives (Ensley, 1995).

However, studies on luggage screening (Rieger et al.,

2021) highlight how factors like time pressure, au-

tomation aids, and target expectancy inﬂuence visual

search behavior and performance in security contexts.

For that reason, new visualization techniques (Klein,

2008; Hurter et al., 2014; Traor

e et al., 2018) need

to be ﬁnd in order to improve efﬁciency and safety

of luggage inspection. According to Milgram, direct

viewing the virtual data allows the user to better un-

derstand the link with the real world (Milgram and

Kishino, 1994). Therefore, in these types of environ-

ments, video mapping technology has the potential to

signiﬁcantly accelerate decision-making processes by

enhancing situational awareness and providing real-

time visual augmentations (Berard and Louis, 2017;

Wang et al., 2020; Sutherland et al., 2019; Douglas

et al., 2016). By projecting images or information di-

rectly onto complex surfaces, video mapping allows

for quick and precise visualization of critical data in

a way that is easily interpretable. For instance, Pana-

sonic’s Medical Imaging Projection System (Nishino

et al., 2018) enables surgeons to perform precise in-

terventions without needing to switch focus between

monitors and the patient, signiﬁcantly improving sur-

gical efﬁciency and decision-making speed during

complex procedures. This approach can be extended

to other high-stakes environments, such as airport se-

curity, where the ability to quickly and accurately vi-

sualize objects can streamline inspections and reduce

decision-making times. Nonetheless, as shown on

Figure 2, projection mapping can distort colors due to

the interaction between the projected light and the sur-

face properties, such as texture, color, and reﬂective

characteristics. These interactions cause color shifts

or accuracy loss in the projection. In critical environ-

ments, this distortion can lead to misleading informa-

tion. Indeed, the operator can misinterpret the ma-

terial information provided by the color and take the

wrong decisions.

This paper aims to be a fundation for larger-scale

studies on color correction algorithm for luggage in-

spection videomapping.

2 RELATED WORK

The issue of radiometric compensation in video map-

ping has been explored through various methods.

A notable contribution comes from Yoshida et al.

Figure 2: Projection on colored surfaces. Green reﬂected on

a red surface becomes orange and blue reﬂected on a green

surface becomes green. In luggage search, this means that,

on the left, a medium-density material (green), like plastic

explosives, would appear the same as a piece of clothe (or-

ange). On the right, a metal (blue), would appear the same

as medium-density materials (green).

(Yoshida et al., 2003), who proposed an approach

based on solving optical equations. Their method

provides precise control of color ﬁdelity across dif-

ferent surfaces by compensating for variations in the

surface’s reﬂection and absorption properties. The

approach works by projecting four reference colors

onto the scene and calculating a transformation matrix

for real-time adjustments. Following the approach by

Yoshida et al., Matthew Post et al. (Post et al., 2018)

and Anselm Grundh

ofer & Daisuke Iwai (Grundh

ofer

and Iwai, 2015) developed two signiﬁcant techniques

aimed at applying colorimetric compensation without

the need for prior radiometric calibration of the video

projector. This advancement reduced the setup time

traditionally required in video mapping applications,

making these techniques more practical in scenarios

where quick adaptations are needed.

However, while these methods are effective in

static environments, they proves less efﬁcient in dy-

namic contexts where surfaces move or change shape,

as the calibration requires projecting these reference

colors whenever the scene changes, limiting its ap-

plicability in scenarios with frequent or rapid change.

In critical and time-sensitive contexts, such as airport

luggage inspections or medical surgeries, delays or

interruptions in the workﬂow can occur. These inter-

ruptions happen because the projection system must

readjust for each new item or surface condition. Thus,

while these methods signiﬁcantly advanced the ﬁeld,

they still do not fully address the challenge of main-

taining continuous, real-time color accuracy with-

out interruptions (Post et al., 2018; Grundh

ofer and

Iwai, 2015). Nevertheless, Anselm Grundh

ofer and

Oliver Bimber advanced colorimetric correction by

developing a technique that adapts to real-time vari-

ations in the scene (Grundh

ofer and Bimber, 2006).

Their method allows the system to continuously ad-

just the projection as the surface properties, such as

texture and lighting, change. This real-time adapt-

ability made their approach highly effective for dy-

IVAPP 2025 - 16th International Conference on Information Visualization Theory and Applications

948

namic environments, ensuring the overall consistency

of the projected image. However, their technique re-

quires certain compromises regarding the brightness

and contrast of the image. These concessions arise be-

cause their primary objective is to maintain the global

integrity of the image rather than ensuring perfect

color ﬁdelity. As a result, while the projection re-

mains stable and adaptable, the precise accuracy of

colors is sometimes sacriﬁced in favor of a broader

focus on maintaining image stability under variable

conditions.

Recent advances (Park et al., 2022; Bokaris et al.,

2015) have explored the use of optimization tech-

niques and artiﬁcial intelligence (AI) for surface

recognition and projection adjustments. These meth-

ods leverage machine learning algorithms to dynami-

cally detect surface properties and adapt the projected

content accordingly, enhancing the system’s ability

to deal with complex and varied surfaces. While

these approaches are promising in terms of increasing

adaptability and reducing the need for manual inter-

vention, they come with signiﬁcant drawbacks. The

most notable issues are visual artifacts and loss of

detail, particularly when ﬁne-grained accuracy is re-

quired. These artifacts can be highly problematic in

contexts where precise inspection of objects is cru-

cial, such as in security screening or medical surg-

eries, where even minor inaccuracies can lead to sig-

niﬁcant errors or oversights. Nonetheless, their work

also highlighted the limitations of Digital Light Pro-

cessing projectors (DLP), which decompose colors

at high speeds, making it difﬁcult to achieve ﬂexible

and precise colorimetric correction in rapidly chang-

ing environments.

These existing techniques highlight both the

progress and limitations in the ﬁeld of color correc-

tion for video mapping. While substantial improve-

ments have been made, existing methods often strug-

gle with accurate color reproduction when project-

ing onto surfaces with varying textures, materials,

and lighting conditions, which is crucial when colors

carry speciﬁc meanings, such as indicating threats in

luggage inspections or guiding surgical interventions.

Our contribution is a novel approach for im-

proving color ﬁdelity in video mapping, particu-

larly in critical environments such as airport secu-

rity and medical settings. To address these chal-

lenges, our approach capitalizes on the speciﬁc con-

textual consistency of the materials used in projec-

tion environments, such as fabric in luggage inspec-

tions. Although the colors of these surfaces may

vary—depending on the objects within a bag—the

underlying material properties remain relatively con-

stant. This consistency allows us to develop a

color correction algorithm based on a pre-constructed

database. This avoids the need for recalibration

for every new surface or lighting change, making

our method usable in critical time-constrained en-

vironments. This method can easily be adapted to

other materials or use cases by simply expanding the

database.

3 MATERIALS AND METHODS

In this paper, we focus on the luggage inspection sce-

nario, where the goal is to identify and differenti-

ate between materials based on their projected col-

ors. In such environments, the surfaces being in-

spected are typically colored fabrics found in lug-

gage—representing various items of clothing, tex-

tiles, and other objects. The challenge is to main-

tain accurate color representation across these diverse

surfaces to properly identify the types of materials

present. To aid in this identiﬁcation process, we

project three distinct colors onto the fabric surfaces,

each representing a different type of material: orange

for organics (e.g., food or leather), green for inorgan-

ics (e.g., plastics), and blue for metals. Therefore,

these are the colors and material categories that will

form the basis of our investigation. Our goal is to en-

sure that, regardless of the color or texture of the fab-

ric surfaces being projected onto, the system can ac-

curately represent these materials using the assigned

colors. Any deviation in color ﬁdelity could result

in misidentifying the nature of the objects in the lug-

gage, affecting both security and efﬁciency.

This section is organized into three parts. The

ﬁrst part details the construction of a colorimetric

database, which serves as the foundation for our al-

gorithm. The subsequent parts describe the algorithm

itself and the two user experiments conducted to vali-

date our hypotheses.

3.1 Construction of the Colorimetric

Database

With this context in mind, we proceed to build a

colorimetric database to support dynamic color cor-

rection during real-time baggage inspections. This

database included three key elements: ID, which

uniquely identiﬁes each surface tested, Projected

Color (the color projected onto the surface), and Per-

ceived Color (the color as it is reﬂected and per-

ceived). We sampled RGB (Red, Green, Blue) values

ranging from 0 to 255, in increments of 5, generat-

ing a total of 52³ elements for each of the three pieces

of fabric. An example is represented on Table 1. To

Colorimetric Compensation in Video Mapping for Luggage Inspection

949

achieve this, as shown on Figure 3 we project the

complete range of colors that the projector can repro-

duce (known as the gamut) onto materials that share

similar textures and properties but differ in color.

Figure 3: Building of the colorimetric database. To build it,

the complete gamut of the projector has been projected onto

three pieces of fabrics.

By projecting the full spectrum of colors, we en-

sure that our database captures the entire set of pos-

sible interactions between the projected light and the

surface, providing a comprehensive reference for real-

time color correction. A lamp-based projector was

used to avoid DLP color deconstruction. The materi-

als chosen for this experiment consist of three pieces

of clothing in the colors red, green, and blue. We

chose them to cover a maximum range of the color

wheel with a minimum of colors. Each garment is

made from the same type of fabric to ensure that

only the color—not the texture or material proper-

ties—varies across the projection surfaces. This al-

lows us to isolate the effect of color on the projection,

ensuring that the colorimetric data recorded reﬂects

the response to different hues rather than surface vari-

ations. We measure the reﬂected color using a cam-

era with manual exposure and white balance settings.

Calibration of the projector and camera is not neces-

sary in this context, as the primary objective is to test

the method itself rather than to fully optimize the sys-

tem for practical applications like baggage inspection.

This approach allows us to focus on the method’s ef-

fectiveness without being constrained by hardware-

speciﬁc adjustments. (Figure 3)

3.2 First Experiment: Testing Color

Corrections for Surfaces in the

Database

The next step in our process is to determine the op-

timal color correction for each surface in the colori-

metric database. Since we have recorded a sample the

gamut of perceptible colors for the three selected sur-

faces, our task now is to select the best matching color

for each surface in real-time projections.

To ensure optimal color ﬁdelity, we tested two

different selections for minimizing the difference be-

tween the expected color (the color intended for pro-

jection) and the reﬂected color (the color perceived

after projection on the surface). The industry standard

for measuring color differences is the DeltaE 2000

(CIEDE2000)(Sharma et al., 2005) in CIELab col-

ors space. We used it for our ﬁrst correction method.

We called this method DeltaE Correction. However,

the presented scenario do not requires color accuracy

in the traditional sense but rather when the color re-

sults in a misclassiﬁcation. Therefore, for our sec-

ond correction method, we used the HSL color space

and we selected the closest color from the colorimet-

ric database with a maximum deviation of 5 degrees

in the hue angle. This method ensures that the pro-

jected color maintains a perceptually close hue to the

intended color, while also ensuring that the saturation

and brightness remain within acceptable thresholds

to ensure the recognition of the hue. We called this

method Hue Correction.

To evaluate the performance of the proposed cor-

rections, we conducted an experiment designed to test

the system’s ability to maintain color ﬁdelity across

various projection surfaces. The experiment involved

projecting different colors onto a range of materials

and asking participants to rank the accuracy of the

color reproduction against a reference color. The ex-

periment involved 17 participants (4 females and 13

males), aged between 25 and 38, all without color

blindness and from diverse research labs. Three types

of fabric were selected as projection surfaces, repre-

senting red, green, and blue base materials. For the

projections, three target colors—orange, green, and

blue—were chosen, along with their respective cor-

rections, as these correspond to key material cate-

gories in baggage inspection: organic, inorganic, and

metallic. During each trial, participants were shown

three variations of the projected colors on the same

surface: one without any color correction, one with

Hue Correction applied, and one with DeltaE Correc-

tion.

As shown on Figure 4, the participants were

tasked with ranking the three projected colors from

closest to farthest in relation to a reference color dis-

played alongside them. This ranking system allowed

us to gather data on which method produced the most

perceptually accurate colors on the different fabrics.

Each participant completed four trials for each color-

surface combination to account for variability in per-

ception and ensure the robustness of the results. This

means that in total, each of the participants completed

36 rankings (four trials for each of the three colors

projected on the three pieces of fabric). On aver-

IVAPP 2025 - 16th International Conference on Information Visualization Theory and Applications

950

Table 1: An example of the colorimetric database, showing each surface’s ID, the projected color, and the perceived reﬂected

color. The data provided a reference for the system to apply color correction.

ID Projected Color (R,G,B) Reﬂected Color (R,G,B)

2 250,100,25 52,97,87

2 255,100,25 52,97,87

2 0,105,25 0,102,91

age, participants completed the task in ﬁfteen min-

utes. Colors and pieces of fabric were presented in a

randomized order.

Figure 4: During the ﬁrst experiment, participants were

asked to rank colors from the closer to the farthest to the

reference color projected above.

The results of this experiment are discussed in de-

tail in the following sections, where we analyze the

effectiveness of each correction method across differ-

ent materials. The generalization of the color correc-

tions to new surfaces has been tested throught a new

experiment.

3.3 Second Experiment: Testing the

Color Corrections on New Surfaces

Since it is impossible to have a database containing

all the possible colors, the purpose of our second ex-

periment is to test the generalization of our correction

methods to new surfaces made of the same material

but different colors. To generalize these two methods,

we implemented an interpolation method, which al-

lows the algorithm to estimate the appropriate color

correction for surfaces that were not part of the initial

database. As the colors have three components, we

needed a three dimensional linear interpolation. Thus,

we decided to use the generalization of Shepard’s in-

terpolation method(Shepard, 1968). This method al-

lowed us to compute a weighted average of the known

color values from the database. Shepard’s method as-

signs greater weight to the colors that are closer to the

current surface’s properties, ensuring that the ﬁnal in-

terpolated color closely matches the surface’s behav-

ior. This minimized the color discrepancy between

the targeted color and the reﬂected color (Figure 5).

To test the generalized corrections, we proceeded

to repeat the previous experiment using new projec-

tion surfaces. Results are discussed in the next sec-

tions.

Figure 5: Pipeline of the color correction algorithm. The

camera-captured color is compared to the database using

Shepard’s interpolation, and the resulting weights are used

to compute the corrected color for projection.

4 RESULTS

This section presents the colorimetric database

recorded to compute the color correction methods and

the results of the two experiments which aimed at test-

ing the methods and their generalization.

4.1 The Colorimetric Database

The ﬁrst phase of the experiment focused on con-

structing the colorimetric database, which served as

the foundation for the color correction process. In

typical color science, a color gamut is often repre-

sented using a chromaticity diagram, such as the CIE

1931 chromaticity diagram, which maps out colors

based on their hue and saturation. This 2D diagram

provides a clear way to visualize the range of colors

that can be produced by a device or material. How-

ever, while these diagrams are widely used, they have

an important limitation: they do not account for lu-

minance, which plays a crucial role in how colors are

perceived.

When representing the gamut reﬂected on pieces

of fabrics in the HSL color space (Hue, Saturation,

Luminance), it becomes evident that although we can

observe all hues, there is a loss of luminance in most

parts of the spectrum (Figure 6). In this space, while

hue and saturation are maintained, the luminance of

colors can vary signiﬁcantly across the same hue,

leading to colors that appear darker or lighter depend-

ing on their positioning in the space. This loss of

brightness can affect the perception of colors, espe-

cially in real-world applications such as video map-

ping, where color ﬁdelity across varying surfaces and

lighting conditions is critical.

Colorimetric Compensation in Video Mapping for Luggage Inspection

951

Figure 6: Representation of the projector’s color gamut in

the HSL space reﬂected on different surfaces. (a) The full

HSL space. (b) The gamut reﬂected on a green fabric sur-

face. (c) The gamut reﬂected on a red fabric surface.

4.2 The First Experiment

To compare the performance of the three color cor-

rection methods—without correction, Hue Correc-

tion, and DeltaE Correction—we applied the Fried-

man test (Sheldon et al., 1996), a non-parametric sta-

tistical test used to detect differences in participants’

rankings across repeated measures.

In order to highlight any differences between how

the corrections performed on various colors, we sepa-

rated the tests for the orange, green, and blue projec-

tions (Figure 7). This allowed us to evaluate whether

the performance of the color correction methods var-

ied depending on the color projected onto the sur-

faces.

Figure 7: Results of the ﬁrst experiment. Both corrections

seem to perform better than the original color for the orange,

but not for the blue. For the green, the differences between

the methods are not signiﬁcant.

• Orange: the results of the Friedman test indicated

a signiﬁcant difference between the three color

correction methods, with a p-value less than 0.01,

demonstrating strong statistical evidence that the

correction methods impacted participants’ per-

ception of color accuracy. The mean ranks of the

three methods were as follows: 1.30 for no cor-

rection, 0.85 for the Hue Correction, and 0.86 for

the DeltaE Correction. These values indicate that

the corrected methods were ranked signiﬁcantly

higher in terms of color accuracy compared to the

uncorrected condition.

• Green: the mean rankings were 0.92 for no cor-

rection, 0.94 for the Hue Correction, and 1.10

for the DeltaE Correction, with p-value less than

0.05, suggesting a moderate difference between

the methods.

• Blue: the Friedman test resulted in a p-value less

than 0.01, indicating a highly signiﬁcant differ-

ence between the methods. The mean rankings

were 0.75 for no correction, and 1.2 for both

the Hue Correction and DeltaE Correction, with

uncorrected condition signiﬁcantly outperforming

both methods.

The results show that the Hue Correction method

performed signiﬁcantly better than the uncorrected

condition for the orange color, where it demonstrated

superior color ﬁdelity. However, this same method

proved ineffective for the blue color. This suggests

that while the Hue Correction excels in certain con-

texts, it struggles with colors like blue, where the

reﬂected color loses too much saturation and lumi-

nance.

4.3 The Second Experiment

The second experiment was conducted with the same

parameters as the ﬁrst one, allowing us to use the

same analytical tools to evaluate the results (Figure 8).

Figure 8: Results of the second experiment. They are sim-

ilar to the previous result, which conﬁrms the efﬁciency of

Shepard’s Distance to compare similar textures.

• Orange: the p-value less than 0.01 indicated a

signiﬁcant difference, with mean rankings of 1.20

for no correction, 0.86 for Hue Correction, and

0.95 for DeltaE, conﬁrming the superiority of the

Hue Correction for maintaining color ﬁdelity.

• Green: the mean rankings were 0.86 for no cor-

rection, 0.95 for Hue Correction, and 1.20 for

DeltaE, suggesting that the corrections were not

able to do better than the original color.

IVAPP 2025 - 16th International Conference on Information Visualization Theory and Applications

952

• Blue: the p-value less than 0.01 indicates an ex-

tremely signiﬁcant difference between the meth-

ods. The mean rankings were 0.65 for no cor-

rection, 1.10 for Hue Correction, and 1.30 for

DeltaE, conﬁrming that the corrections methods

struggle to accurately reproduce blue hues.

The use of Shepard’s interpolation to compare

similar projection surfaces appears to be effective,

as evidenced by the results of the second experi-

ment, which closely align with the ﬁndings from the

ﬁrst. The consistency in performance across both

experiments suggests that this approach successfully

maintains color ﬁdelity across different projection

surfaces. The results were similar, conﬁrming that

the Hue Correction method worked well for orange,

while the uncorrected provided better results for blue

hue.

5 DISCUSSION

In this paper, we conducted a series of experiments

to have a fundation for large-scale study on our color

correction algorithm. Using a colorimetric database

and Shepard’s interpolation, we compensated color

variations across different surfaces with similar tex-

tures in video mapping. The goal was to maintain

color ﬁdelity in real-time projections without frequent

recalibration, ensuring smooth and accurate color re-

production across varied projection surfaces. The re-

sults demonstrate that our color correction method

signiﬁcantly outperforms the uncorrected condition

for colors close to orange. Within this range, the

hue of the light tends to shift slightly upon reﬂection,

which is why the algorithm can effectively compen-

sate for these deviations.

However, both correction methods struggled with

blue hues, as indicated by consistently poor results in

both trials. This result could have been predicted from

the database as the Figure 6 already showed that sur-

faces tend to do not reﬂect blue hues. This suggests

that the reﬂected blue color experiences a signiﬁcant

loss in luminance and saturation, making it difﬁcult

for the human eye to correctly identify the hue. The

loss of brightness causes the blue to appear too dark,

while the reduction in saturation diminishes its vivid-

ness, leading to a misinterpretation of the color. This

highlights a fundamental limitation of projectors, par-

ticularly when handling colors like blue. The pro-

jector’s inability to maintain sufﬁcient luminance and

saturation in the reﬂected light signiﬁcantly hampers

the accurate reproduction of blue hues.

One potential solution to address the limitations

of projector color accuracy, particularly for blue hues,

could be to explore the concept of simultaneous con-

trast(Mittelst

adt et al., 2014). This phenomenon oc-

curs when the perception of a color is inﬂuenced

by the surrounding colors, potentially enhancing the

brightness and saturation of colors that might other-

wise appear too dark or muted. By adjusting the con-

trast of the surrounding colors in the projected scene,

it may be possible to counterbalance the loss of lumi-

nance and saturation in the blue hues, making them

appear more vivid and recognizable. Integrating si-

multaneous contrast adjustments into the color cor-

rection algorithm could help overcome the projector’s

inherent limitations (Akiyama et al., 2018). However,

this approach would require modifying the colors in

the projected image. In critical environments, where

the accurate representation of colors plays a crucial

role in decision-making, altering the surrounding col-

ors to improve the perception of blue could potentially

compromise the clarity or meaning of the projected

information. Therefore, while simultaneous contrast

is a promising solution, further research is necessary

to explore its applicability in contexts where color ﬁ-

delity is essential and must remain consistent.

The algorithm could be easily adapted to other en-

vironments by adding additional projection surfaces

to the colorimetric database. As more surfaces are in-

corporated, the system becomes capable of handling a

wider range of materials and lighting conditions, fur-

ther enhancing its versatility. This ﬂexibility allows

the algorithm to remain effective in environments be-

yond those initially tested. This makes the algorithm

suitable for critical environments where hues within

this spectrum are crucial for decision-making.

Being able to project any desired color not only

enhances color ﬁdelity of the interface, but it also

allows for dynamic alterations in the appearance of

physical objects (Amano et al., 2012; Iwai and Sato,

2011). This capability opens up the potential for new

interactions by changing how objects are perceived.

For instance, it would be possible to diminish reality

(Mori et al., 2017) to help the user focusing on his

task.

ACKNOWLEDGEMENTS

We acknowledge the use of GPT-4o to correct the

grammar and syntax.

REFERENCES

Akiyama, R., Yamamoto, G., Amano, T., Taketomi,

T., Plopski, A., Sandor, C., and Kato, H. (2018).

Colorimetric Compensation in Video Mapping for Luggage Inspection

953

Light projection-induced illusion for controlling ob-

ject color. In 2018 IEEE Conference on Virtual Reality

and 3D User Interfaces (VR), pages 499–500. IEEE.

Amano, T., Komura, K., Sasabuchi, T., Nakano, S., and

Yamashita, S. (2012). Appearance control for human

material perception manipulation. In Proceedings of

the 21st international conference on pattern recogni-

tion (ICPR2012), pages 13–16. IEEE.

Berard, F. and Louis, T. (2017). The object inside: As-

sessing 3d examination with a spherical handheld

perspective-corrected display. In Proceedings of the

2017 CHI Conference on Human Factors in Comput-

ing Systems, pages 4396–4404.

Bertin, J. (1983). Semiology of graphics.

Bokaris, P.-A., Gouiff

es, M., Jacquemin, C., Chomaz, J.-

M., and Tr

emeau, A. (2015). One-frame delay for

dynamic photometric compensation in a projector-

camera system. In 2015 IEEE International Confer-

ence on Image Processing (ICIP), pages 2675–2679.

IEEE.

Douglas, D. B., Petricoin, E. F., Liotta, L., and Wilson, E.

(2016). D3d augmented reality imaging system: proof

of concept in mammography. Medical Devices: Evi-

dence and Research, pages 277–283.

Drebin, R. A., Carpenter, L., and Hanrahan, P. (1988). Vol-

ume rendering. ACM Siggraph Computer Graphics,

22(4):65–74.

Ensley, M. (1995). Toward a theory of situation awareness

in dynamic systems. Human factors, 37:85–104.

Gibson, E. J. and Pick, A. D. (2000). An ecological ap-

proach to perceptual learning and development. Ox-

ford University Press, USA.

Grundh

ofer, A. and Bimber, O. (2006). Real-time adaptive

radiometric compensation. In ACM SIGGRAPH 2006

Research posters, pages 56–es.

Grundh

ofer, A. and Iwai, D. (2015). Robust, error-tolerant

photometric projector compensation. IEEE Transac-

tions on Image Processing, 24(12):5086–5099.

Hurter, C., Taylor, R., Carpendale, S., and Telea, A. (2014).

Color tunneling: Interactive exploration and selection

in volumetric datasets. In 2014 IEEE Paciﬁc Visual-

ization Symposium, pages 225–232. IEEE.

Iwai, D. and Sato, K. (2011). Document search support by

making physical documents transparent in projection-

based mixed reality. Virtual reality, 15:147–160.

Kindlmann, G. (2002). Transfer functions in direct vol-

ume rendering: Design, interface, interaction. Course

notes of ACM SIGGRAPH, 3.

Klein, G. (2008). Naturalistic decision making. Human

factors, 50(3):456–460.

Metz, C. E. and Doi, K. (1979). Transfer function analysis

of radiographic imaging systems. Physics in Medicine

& Biology, 24(6):1079.

Milgram, P. and Kishino, F. (1994). A taxonomy of mixed

reality visual displays. IEICE TRANSACTIONS on In-

formation and Systems, 77(12):1321–1329.

Mittelst

adt, S., Stoffel, A., and Keim, D. A. (2014). Meth-

ods for compensating contrast effects in information

visualization. In Computer Graphics Forum, vol-

ume 33, pages 231–240. Wiley Online Library.

Mori, S., Ikeda, S., and Saito, H. (2017). A survey of di-

minished reality: Techniques for visually concealing,

eliminating, and seeing through real objects. IPSJ

Transactions on Computer Vision and Applications,

9:1–14.

Nishino, H., Hatano, E., Seo, S., Nitta, T., Saito, T., Naka-

mura, M., Hattori, K., Takatani, M., Fuji, H., Taura,

K., et al. (2018). Real-time navigation for liver surgery

using projection mapping with indocyanine green ﬂu-

orescence: development of the novel medical imaging

projection system. Annals of surgery, 267(6):1134–

1140.

Park, J., Jung, D., and Moon, B. (2022). Projector com-

pensation framework using differentiable rendering.

IEEE Access, 10:44461–44470.

Post, M., Fieguth, P., Naiel, M. A., Azimifar, Z., and Lamm,

M. (2018). Fast radiometric compensation for nonlin-

ear projectors. Journal of Computational Vision and

Imaging Systems, 4(1):3–3.

Rieger, T., Heilmann, L., and Manzey, D. (2021). Visual

search behavior and performance in luggage screen-

ing: effects of time pressure, automation aid, and tar-

get expectancy. Cognitive Research: Principles and

Implications, 6:1–12.

Sharma, G., Wu, W., and Dalal, E. N. (2005). The

ciede2000 color-difference formula: Implementation

notes, supplementary test data, and mathematical ob-

servations. Color Research & Application, 30(1):21–

30.

Sheldon, M. R., Fillyaw, M. J., and Thompson, W. D.

(1996). The use and interpretation of the friedman test

in the analysis of ordinal-scale data in repeated mea-

sures designs. Physiotherapy Research International,

1(4):221–228.

Shepard, D. (1968). A two-dimensional interpolation func-

tion for irregularly-spaced data. In Proceedings of the

1968 23rd ACM national conference, pages 517–524.

Sutherland, J., Belec, J., Sheikh, A., Chepelev, L., Al-

thobaity, W., Chow, B. J., Mitsouras, D., Christensen,

A., Rybicki, F. J., and La Russa, D. J. (2019). Apply-

ing modern virtual and augmented reality technolo-

gies to medical images and models. Journal of digital

imaging, 32:38–53.

Traor

e, M., Hurter, C., and Telea, A. (2018). Interactive

obstruction-free lensing for volumetric data visualiza-

tion. IEEE transactions on visualization and computer

graphics, 25(1):1029–1039.

Treisman, A. M. and Gelade, G. (1980). A feature-

integration theory of attention. Cognitive psychology,

12(1):97–136.

Wang, X., Besanc¸on, L., Rousseau, D., Sereno, M., Ammi,

M., and Isenberg, T. (2020). Towards an understand-

ing of augmented reality extensions for existing 3d

data analysis tools. In Proceedings of the 2020 CHI

Conference on Human Factors in Computing Systems,

pages 1–13.

Yoshida, T., Horii, C., and Sato, K. (2003). A virtual color

reconstruction system for real heritage with light pro-

jection. In Proceedings of VSMM, volume 3, pages

1–7. Citeseer.

IVAPP 2025 - 16th International Conference on Information Visualization Theory and Applications

954