Experimental Evaluation of Camouflage Effectiveness Against

Ground-Based Surveillance

Viktor Vitoul

, Jan Ivan

, Ladislav Potužák

Michal Šustr

and Barbora Hanková

Fire Support Department, University of Defence, Kounicova 65, Brno, Czech Republic

Keywords: Perception and Awareness, Thermal Imagining, Optical Sensors, Artillery Multidomain Operation,

Artillery Joint Fire Support, Military Camouflage, Military Equipment, Military Deception, Field Experiment,

Mechanical Sensors, Force and Tactile Sensors, Artillery, Image Processing.

Abstract: Camouflaging mortar firing positions represents a critical force protection measure in modern conflicts,

aiming to prevent enemy observation and subsequent destruction. The objective of this pilot study is to

evaluate the effectiveness of various camouflage techniques in concealing mortars, ammunition assets, and

support equipment from detection by selected ground-based reconnaissance means. The experimental phase

employed a range of artillery reconnaissance sensors, optical devices, and unaided visual observation. The

observed targets including mortar firing positions of various calibres and decoy positions were camouflaged

using different methods and levels of concealment, and deployed in terrain with varying vegetation density

and spatial characteristics. The detected differences in target visibility highlight the strengths and limitations

of individual observation methods depending on target characteristics and environmental conditions. The

findings of this pilot study offer practical recommendations for the effective camouflage of mortar units in

current operational environments.

1 INTRODUCTION

The current battlespace of multi-domain operations,

characterized inter alia by high-intensity sensor

surveillance and widespread deployment of

unmanned systems, imposes extraordinary demands

on unit protection—particularly in terms of

camouflage. Mortar units are especially vulnerable to

detection by enemy artillery reconnaissance assets,

notably during emplacement in firing positions and

while conducting fire missions (Sedláček et al., 2023;

Havlík et al. 2022). This vulnerability underscores the

necessity for rational planning and allocation of

defence resources under conditions of increasing risk

(Šlouf et al., 2023).

Each artillery discharge generates a visual,

acoustic, and thermal signature. The ongoing war in

Ukraine has demonstrated that the integration of

unmanned aerial vehicles (UAVs), thermal imaging

https://orcid.org/0009-0000-3624-7986

https://orcid.org/0000-0002-6194-8482

https://orcid.org/0000-0002-0213-717X

https://orcid.org/0000-0002-7342-7641

https://orcid.org/0009-0008-3444-591X

devices, and multi-sensor platforms with an efficient

kill chain can significantly increase the probability of

early target acquisition and

rapid destruction of firing

positions (Khoma, 2023; Ali et al., 2023). As a result,

camouflage has become one of the critical

components of artillery force protection.

With the ongoing advancement of technologies,

sophisticated command and control systems

(C2/C4ISR) are increasingly coming to the forefront.

These systems enable the real-time integration of

sensor data and significantly accelerate the decision-

making cycle (Amphenol Aerospace, 2023). Within

this framework, artificial intelligence (AI) is

beginning to assert its role, facilitating automated

object detection, behaviour prediction, and enhancing

the probability of identifying concealed targets

through machine learning methods (Beals, 2023).

Traditional camouflage techniques, when not adapted

to address these emerging threats, can be easily

Vitoul, V., Ivan, J., Potužák, L., Šustr, M. and Hanková, B.

Experimental Evaluation of Camouﬂage Effectiveness Against Ground-Based Surveillance.

DOI: 10.5220/0013712500003982

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

In Proceedings of the 22nd Inter national Conference on Informatics in Control, Automation and Robotics (ICINCO 2025) - Volume 1, pages 425-434

ISBN: 978-989-758-770-2; ISSN: 2184-2809

425

defeated in combination with image recognition

algorithms (Ali et al., 2023). Therefore, future

development of camouflage systems must be closely

aligned with advancements in AI, computer vision,

and predictive analytics, in order to maintain

operational effectiveness even in environments

characterized by intense digital reconnaissance

(Hughes, 2024). In future implementations,

integrating artificial intelligence and computer vision

into the decision-making process may significantly

enhance the reliability of camouflage assessment and

deployment, particularly by mitigating human error

and enabling automated terrain analysis from UAV

imagery.

Effective protection and concealment of artillery

units require a multi-faceted approach (Ivan et al.,

2022). Currently employed methods include natural

camouflage, technical camouflage systems (e.g.,

camouflage nets), the deployment of decoy firing

positions, and the implementation of "shoot and

scoot" tactics. The overarching objective of these

measures is to reduce the likelihood of unit detection

by enemy reconnaissance assets.

Minimization of visual reflections, suppression of

thermal and electromagnetic signatures, and

environmental adaptation of deployed units also play

a critical role in modern camouflage practices

(Khoma, 2023; Zhang et al., 2021). In contemporary

armed forces, camouflage often relies on traditional

methods (e.g., disruptive patterns, natural materials)

while simultaneously undergoing modernization. For

example, the introduction of the MAD21 camouflage

pattern is intended to enhance concealment in both

natural and urban environments (CZ Defence, 2022).

Parallel to these operational measures, scientific

methodologies have emerged to quantify camouflage

effectiveness. Notable research directions include the

evaluation of visual similarity between objects and

their surroundings based on human perception (Li et

al., 2022), the application of network analysis and

clustering algorithms (Kim, Yang & Kwon, 2021),

and the development of adaptive materials aimed at

reducing thermal signatures (Su & Zhao, 2023).

Advances in detection capabilities have been

driven by the evolution of deep neural networks

utilizing datasets such as MCAM (Hwang & Ma,

2024), and model architectures like MilDETR (Li et

al., 2024), CAMOUFLAGE-Net (Karthiga &

Asuntha, 2025), and YOLOv5/YOLOv7, which have

proven effective when deployed on UAV platforms

(Zeng et al., 2024).

Further studies have examined the impact of

visualization techniques on human decision-making,

particularly the risk of situational awareness

degradation due to excessive target highlighting.

When digital interfaces overemphasize objects

through colour, size, or motion observers may focus

disproportionately on marked elements and overlook

other critical aspects of the tactical environment

(Gardony et al., 2022).

Some lines of research draw inspiration from

biological mechanisms, such as mimicry, bionics, and

adaptive behaviour, which are being applied in the

development of next-generation camouflage systems

(Matthews et al., 2024). However, most of these

studies are conducted under laboratory or simulated

conditions. There remains a lack of experimentally

grounded research that evaluates the effectiveness of

camouflage techniques in real terrain settings and

against a broad spectrum of modern sensors

conditions that closely approximate the operational

battlefield.

This pilot study aims to address this research gap.

The primary objective is to assess the effectiveness of

selected camouflage and deception methods applied

to mortar units under conditions resembling current

combat environments. The study seeks to contribute

to a more comprehensive understanding of protection

strategies for artillery firing positions. At its core, this

research is based on a field experiment.

Based on the outlined context, the study was

designed to address the following research questions

(RQ):

RQ1: What is the difference in detection rates of

camouflaged positions depending on the applied

camouflage method?

RQ2: Which type of sensor is most successful in

detecting realistically camouflaged targets?

RQ3: Can decoy positions be effectively used as

part of a tactical approach to reduce detection

probability?

Drawing from existing knowledge, the working

hypothesis posits that natural camouflage will be

more effective than decoy positions in open terrain

when observed through optical means, while

camouflage nets are expected to perform less

effectively in the infrared spectrum. To test this

assumption, a null hypothesis (H₀) was formulated,

stating that there is no statistically significant

difference in detectability among the different types

of camouflage.

2 EXPERIMENTAL

FRAMEWORKS

This research responds to the current operational need

ICINCO 2025 - 22nd International Conference on Informatics in Control, Automation and Robotics

426

to verify the effectiveness of camouflage methods

under conditions that reflect the realities of today’s

sensor-saturated battlefield. Given the widespread

deployment of UAVs, thermal imaging, and

multisensory platforms that enable precise target

acquisition, it is essential to test whether traditional or

modernized camouflage techniques can effectively

reduce the likelihood of detection.

The experiment was designed as an initial pilot

phase of a broader research effort, aimed at

comparing the detectability of objects camouflaged

using various techniques within a selected operational

environment. In addition to measuring direct

detection rates, the study also assessed a range of

supplementary variables relevant to the operational

applicability of sensor systems:

Level of Confidence in Target Detection – a

subjective observer rating on a scale from 1 to 5

indicating the degree of certainty regarding the

correctness of target identification,

Subjective Evaluation of the Observation

System – including clarity of the image, ease of use,

and the observer’s perceived ability to distinguish

between targets,

Type of Detection Error – categorized as Type I

errors (false positives) or Type II errors (missed

detections of actual targets),

Observer Selection Behaviour – the order in

which participants chose specific sensor platforms for

target detection tasks.

The results of this pilot study are intended to

inform the design of a larger-scale experiment

focused on validating camouflage effectiveness

across different seasonal conditions and with an

expanded participant base. This step is essential for

generating repeatable data applicable to the

protection of mortar units operating in contemporary

combat environments.

3 MATERIALS AND METHODS

The primary objective of this study was to

experimentally verify the effectiveness of various

forms of camouflage for mortar firing positions under

realistic field conditions and to analyse their

resistance to detection by ground-based sensor

systems. A secondary objective was to record

variations in the detection rates of individual targets,

including the occurrence of Type I errors (false

positives) and Type II errors (missed detections), as

well as to assess the observers’ subjective confidence

in their detections when employing different types of

camouflage.

3.1 Materials

The parameters for the field experiment were derived

from actual equipment and structures commonly

associated with the operations of mortar units.

Emphasis was placed on creating a diverse set of

observed targets and camouflage configurations,

reflecting both standardized and improvised

concealment techniques typically employed during

real-world deployments.

3.1.1 Variables

The independent variable was the type of

camouflage applied to the target, categorized as

natural camouflage, camouflage netting, no

camouflage, and decoy position.

The dependent variables were: the detection rate

(expressed as a percentage), and the time required to

detect the target.

The controlled variables included: the distance

between the target and the observer, the time of day,

ambient light conditions, and the type of observation

post employed.

3.1.2 Sample Selection

The observation tasks were conducted by students

and members of the Czech Armed Forces (CAF)

representing various military specializations. A total

of 23 students from the Faculty of Military

Leadership participated in the experiment. A

purposive sampling strategy was employed to obtain

a representative sample of users familiar with real

military equipment and operational procedures. The

participants had no prior knowledge of the location or

type of camouflaged and decoy objects.

3.1.3 Environment and Conditions

The experiment was conducted on a designated

training range featuring natural vegetation with

gradual transition into light forest terrain.

Observations were carried out at varying distances

ranging from 150 to 400 meters, allowing the

simulation of dynamic battlefield conditions.

3.2 Methods

Data were recorded using standardized observation

forms, which included:

 Fields for the type of detected object, the

observation system employed;

 The observer’s confidence level (rated on a 1–

5 scale);

Experimental Evaluation of Camouﬂage Effectiveness Against Ground-Based Surveillance

427

 The type of detection error (Type I – false

positive; Type II – omission of an actual

target);

 The estimated distance to the target.

The forms were collected after each observation

block, digitized, and subsequently anonymized for

statistical processing.

For the statistical analysis, a combination of

descriptive and inferential statistical methods was

employed. Descriptive statistics included the

calculation of arithmetic means, standard deviations,

and variances for the observed variables.

To test for statistical associations, a chi-square

test was used to evaluate the relationship between the

type of camouflage and the detection rate.

Furthermore, a one-way analysis of variance

(ANOVA) was conducted to compare the

effectiveness of the different observation systems

employed in the experiment.

3.2.1 Observed Objects and Camouflage

Techniques

Figure 1: Tactical diagram of object positions with distance

scale within the experimental area.

A total of six objects representing either real or decoy

military targets were deployed within the observation

area, as illustrated in Figure 1. Each object employed

a different camouflage method:

1. Decoy light mortar position (81mm),

constructed using a black plastic tube simulating a

mortar barrel, with dimensions corresponding to a

real weapon system. The object was placed

approximately 180 meters from the observation post.

2. Command-type Vehicle, concealed under a

camouflage net, positioned at the edge of a forested

area. The vehicle represented a typical light tactical

automobile. The distance from the observer was

approximately 230 meters.

3. Supplementary Material commonly used at

firing positions, including ammunition crates and

packaging containers for transport and storage

purposes.

4. Medium Mortar Position (120mm), located

in a transition zone between forest and open terrain,

with a single crew member kneeling next to the

weapon system. The observation distance was

approximately 250 meters.

5. Camouflaged Medium Mortar Position

(120mm) in open terrain, utilizing natural vegetation

(grass cover) for concealment. The object was

situated at a distance of approximately 310 meters

from the observation point.

6. Decoy Medium Mortar Position, constructed

from natural materials (logs) cut and arranged to

simulate the dimensions of a real mortar system. The

decoy was placed at approximately 360 meters from

the observer.

An example of natural camouflage and the use of

a technical camouflage system is shown in Figure 2,

which depicts a camouflaged mortar concealed with

natural vegetation and a command-type vehicle

covered with a camouflage net.

Figure 2: Observed objects camouflaged using natural

materials and camouflage net.

3.2.2 Classification and Description of

Sensors

Due to the rapid advancement of technologies,

detection methods used for identifying and evaluating

the effectiveness of camouflage are continuously

evolving. The selection of detection platforms is

therefore a critical factor in the interpretation of results.

Findings from this study may serve as a foundation for

future improvements in military tactics and enhanced

protection against modern reconnaissance systems.

The following detection systems were employed

ICINCO 2025 - 22nd International Conference on Informatics in Control, Automation and Robotics

428

in the experiment, representing a spectrum of

platforms commonly available within the armament

of contemporary armed forces:

Human Eye – direct visual observation without

any optical support.

Aiming Circle (PAB-2A) – featuring

8× magnification and a 5° field of view; used for

general observation and orientation.

Standard Military Binoculars – 7×

magnification with a 7.5° field of view; suitable for

rapid terrain scanning and identification of large

targets.

Spotting Scope Meopta MeoStar S2 82 HD is

high-magnification (up to 70) optical device designed

for detailed observation of distant targets.

Thermal Scope HIKMICRO Falcon FH25 –

passive sensor for detecting heat signatures, enabling

observation regardless of lighting conditions. The is a

compact thermal monocular with a 384 × 288 px, 12

µm VOx sensor and thermal sensitivity of ≤ 20 mK.

It features a 25 mm lens, 8× digital zoom, and a

detection range of up to 1200 meters.

Multisensory Device MOSKITO – combines a

daylight optical channel (8× magnification, ~6° field

of view) and a thermal imaging sensor. In this

experiment, optical channel was used for visual target

identification.

Unmanned Aerial Vehicle (UAV) DJI Mavic 3

– equipped with a wide-angle camera and thermal

sensor; manually controlled by the operator without

automated data processing.

The human eye, as the most fundamental means

of sensory perception, was employed as the primary

tool for initial orientation and target acquisition.

Despite its limited range and dependence on ambient

lighting conditions, it remains a widely used method

for rapid detection of targets in natural terrain.

Subsequently, an aiming circle was utilized. This

instrument not only supports observation but also

enables azimuth determination. While it plays an

important role in the field orientation of artillery

units, it was less frequently used by participants

during the experiment.

A multisensory system combining optical and

thermal imaging capabilities provided a

comprehensive visual and thermal representation of the

target, including distance estimation. Its functionality

under reduced visibility and overall versatility made it

one of the most preferred detection tools.

For spatial situational awareness, an UAV

equipped with a camera and thermal sensor was

employed. The UAV was manually piloted by the

evaluator and enabled elevated-area observation

based on operator discretion. All flights were

conducted during daylight hours.

For detailed long-range observation, a spotting

scope with high optical magnification was used. It

enabled the identification of fine details that were not

easily discernible with standard optics, including

specific camouflage patterns and material structures.

A thermal scope was included in the sensor suite,

offering the advantage of detection based on thermal

radiation. It proved effective for revealing objects

concealed by vegetation or camouflage materials

especially targets such a humans and vehicles. Its

independence from ambient light conditions was a

significant benefit under reduced visibility.

The last tool included was a standard military

binocular, which was frequently used by

participants. It enabled the successful detection of

even well-concealed objects, including a vehicle

hidden at the forest edge.

3.2.3 Validity

To ensure repeatability and scientific validity, the

experiment was designed in such a way that it could

be replicated by other research teams under

comparable conditions. All scenarios, environments,

and configurations were meticulously documented.

Internal validity was supported through the use of a

controlled testing environment and standardized

instructions provided to all participants. External

validity was ensured by selecting test parameters that

closely reflect real-world operational deployment

scenarios of mortar units.

3.2.4 Observation Procedure and Applied

Methods

Each participant began the observation phase using

unaided vision, without any technical support.

Subsequently, participants were allowed to choose

the order in which to use the remaining observation

tools based on personal preference.

A maximum time limit of three minutes was allocated

for the use of each tool to ensure uniform exposure

and consistent use of observation time.

After using each detection system, participants

recorded their observations in a standardized

protocol, which included the following data fields:

 Type of identified target (e.g., firing position,

vehicle, decoy, personnel, equipment);

 Observation tool used;

 Estimated distance to the target;

 Confidence level in detection, on a 1–5 scale.

Experimental Evaluation of Camouﬂage Effectiveness Against Ground-Based Surveillance

429

The results indicate that participants likely

selected observation tools based on familiarity and

perceived ease of use:

 Aiming circle was frequently chosen as the

second tool, possibly due to its simplicity of

operation;

 Spotting scope dominated the third round of

selections, reflecting its strength in optical

magnification;

 Multifunctional sensor systems were more

often used later in the process, suggesting a

preference for simpler tools in the early

evaluation stages;

 Thermal scope was selected in the later

phases, likely due to their specialized

capability in detecting heat signatures;

 UAV (drone) was most commonly deployed

last, likely for targeted final-stage

reconnaissance.

These conclusions are based on observed

selection order, not on direct participant feedback. A

comparison of detection success rates, error types,

and subjective evaluations of each observation

system is presented in Table 1.

Table 1: Sensor Preference Ranking Based on Participant

Selection.

Selection Order

Most

Frequently

Selected Tool

Percentage of

Respondents (%)

2nd Selection aiming circle 35

3rd Selection spotting scope 30

4th Selection MOSKITO 45

5th Selection thermal scope 30

6th Selection UAV 25

The table provides a comprehensive comparative

summary of individual observation systems in terms

of detection success, error rates, and subjective

reliability assessments. For each system, the table

presents the percentage of successfully detected

targets, the number of Type I errors (false detections),

Type II errors (missed actual targets), and the average

subjective confidence rating on a 1 to 5 scale.

The results reveal significant differences between

the systems - some exhibit high detection capability

and low error rates, while others are less reliable or

more demanding to operate. The table thus offers an

integrated comparison of both the technical

effectiveness and the practical usability of each

sensor under the conditions of a field experiment.

3.2.5 Detection Assessment

The evaluation of the detection capabilities of

individual systems was based on three key variables:

detection success (Yes/No), subjective confidence

level, and the classification of detection errors -

specifically, Type I errors (false identification of a

non-existent object) and Type II errors (failure to

detect an actually present target). Each entry in the

observation protocol was further analysed with

respect to the type of sensor used, the identified object

category, and the estimated distance to the target.

For comparison purposes, the data were

aggregated into a summary matrix tracking the

performance of each observation system. The matrix

includes:

 Percentage of correctly detected targets;

 Average subjective confidence rating (scale

1–5);

 Frequency of Type I errors (false positives)

and Type II errors (missed detections);

 Detection success rates across different target

types (e.g., actual firing positions, decoy

targets, camouflaged objects).

Table 2: Results of the comparison of individual

observation tools in terms of detection success, error rates,

and subjective reliability assessment.

Observation

Tool

Detection

Success

Rate (%)

Type I

Errors

Type

Errors

Subjective

Reliability

(1–5)

Human Eye 92 5 3 5

Aiming

circle

75 10 15 3

Spotting

Scope

80 8 12 4

MOSKITO 88 6 6 5

Thermal

scope

70 12 18 2

UAV 65 15 20 2

Binoculars 78 9 13 3

The results are evaluated using descriptive

statistics, specifically arithmetic mean and standard

deviation, and are compared both across observation

tools and across target types. This approach allows for

the identification not only of the overall effectiveness

of each system but also its limitations in specific

operational scenarios, such as lower confidence levels

when detecting naturally camouflaged targets or

higher false positive rates with decoy objects.

The analysis thus provides a quantitative basis for

comparing the detection efficiency of individual

sensors and enables an evaluation of the effectiveness

ICINCO 2025 - 22nd International Conference on Informatics in Control, Automation and Robotics

430

of different camouflage techniques from the

perspective of realistic operational application. A

summary of these key indicators is presented in Table

2, which facilitates a comprehensive comparison of

the observation systems in terms of reliability,

accuracy, and practical usability under field

conditions.

The table provides a comparative summary of

observation tools based on detection success, error

incidence, and subjective reliability ratings. For each

system, it reports the percentage of successfully

detected targets, the number of Type I errors (false

detections), Type II errors (missed actual targets), and

the average subjective rating on a 1 to 5 scale. The

results show clear distinctions between the tools -

some demonstrate high detection performance and

low error rates, while others are less reliable or more

demanding to operate.

This comparison enables a holistic evaluation of

both the technical efficiency and field-level

applicability of each sensor type in a live operational

test environment. While this study focused primarily

on detection success, future research will aim to

quantify additional camouflage performance

indicators, such as time-to-detection and targeting

accuracy, and assess their statistical correlation with

specific camouflage parameters.

4 RESULTS

Descriptive and inferential statistical methods were

employed to process the collected data. Descriptive

statistics were used to evaluate detection rates, error

types, and the average subjective confidence levels

reported by participants. To assess the statistical

significance of relationships between variables, the

chi-square test and one-way analysis of variance

(ANOVA) were applied. The analysis of results

considers both quantitative indicators and the

interrelation between the type of camouflage, the

observation device used, and the success rate of target

detection.

As illustrated in Figure 3, the most frequently

detected object across nearly all observational

methods was the unconcealed mortar, confirming that

the absence of camouflage significantly increases the

probability of detection, regardless of the observation

technology employed. The most effective tool for

visual detection was revealed to be the human eye,

which demonstrated the ability to rapidly and

accurately identify visible objects in most cases.

Notably, its performance surpassed that of certain

technical devices, highlighting the importance of field

experience and innate perceptual acuity in operational

environments.

In contrast, camouflaged objects, particularly

decoy positions, proved significantly more difficult

for observers to detect. The highest detection success

for such targets was achieved using the MOSKITO

device, equipped with multisensory capabilities;

however, even in this case, full and accurate

identification was not consistently ensured. These

results confirm that deception and natural camouflage

techniques substantially reduce the probability of

detection, especially when properly implemented.

Figure 3: Detectability of individual targets based on the

type of reconnaissance asset.

Table 3: Number of successful detections of individual

targets by various reconnaissance assets.

Participant preferences regarding the selection of

individual observation tools are clearly illustrated in

Table 3, which displays the sequence in which tools

were used during the observation trials. The spotting

scope demonstrated consistent, relatively low-

selectivity performance. The number of detections

was evenly distributed across different object types.

This indicates its limited ability to distinguish

between targets under specific conditions and

suggests its suitability more for general observation

rather than the targeted identification of concealed or

decoy assets.

As part of the evaluation, participants’

preferences in the sequence of observation tool usage

were also monitored. Each participant began with

unaided visual observation, followed by optional

selection of additional tools according to personal

Experimental Evaluation of Camouﬂage Effectiveness Against Ground-Based Surveillance

431

judgment. The table below presents the most

frequently chosen tool at each position in the

selection order (e.g., second, third, etc.). The values

indicate the percentage representation of each tool’s

selection at specific positions and serve primarily to

interpret user decision-making behavior, rather than

to evaluate tool effectiveness. An overview of these

selection preferences is provided in Figure 4, which

illustrates the distribution of tools according to their

order of use.

Figure 4: Comparison of the effectiveness of different

camouflage types against various reconnaissance assets.

The data indicate that natural camouflage and

decoy positions achieved the highest levels of

concealment effectiveness, with strong performance

across most observational tools. For instance, when

using the naked eye, natural camouflage yielded

nearly 90% concealment effectiveness, whereas an

unconcealed mortar was detected in 100% of cases,

underscoring the critical role of camouflage

measures. Table 4 shows the percentage differences

in detection success among observation tools across

various target types, including those masked with

natural materials or presented as decoys.

The MOSKITO multisensory device

demonstrated high effectiveness in detecting decoy

positions, but its performance decreased when

identifying naturally camouflaged objects.

Conversely, thermal imaging maintained relatively

consistent results, although camouflage effectiveness

was often lower, particularly for targets with more

prominent thermal signatures.

An important observation is that certain tools such

as the aiming circle showed lower success in

detecting camouflaged or decoy targets, where

concealment effectiveness reached up to 90%. This

finding highlights the limitations of traditional optical

instruments with narrow fields of view, which may

not be suitable for identifying concealed objects in

complex terrain.

Table 4: Detection rate and effectiveness of different

camouflage types by reconnaissance system type.

5 DISCUSSION

Although valuable insights were obtained, several

limitations must be discussed, as they may influence

the results and should be considered in their

interpretation.

The experimental group consisted of 23 students

with varying levels of familiarity and no standardized

experience profile in the use of military optics. This

variability may have affected individual target

recognition performance and represents

a methodological limitation of the study.

Another variable factor potentially affecting the

results was the environmental condition and season in

which the experiment was conducted. Testing took

place in February during the winter when vegetation

offered limited opportunities for natural concealment.

The low density of vegetation increased the visibility

of objects that, during other seasons such as summer,

would typically be more difficult to detect due to

denser foliage and different environmental color

spectra. This factor may have influenced the

effectiveness of natural camouflage and the ability to

detect it. For this reason, it is recommended that the

experiment be repeated. Conducting the study in

summer conditions would enable a comparison with

the winter variant and provide a more comprehensive

understanding of camouflage effectiveness

throughout the year.

The experiment was conducted as a single session

under clear weather conditions, without precipitation,

ensuring favorable lighting. Therefore, the results

cannot be directly extrapolated to low-visibility

scenarios, such as those involving rain, fog, or low-

light conditions. Varying meteorological

circumstances may influence both target visibility

and camouflage effectiveness, and repetition of the

ICINCO 2025 - 22nd International Conference on Informatics in Control, Automation and Robotics

432

experiment under different climatic conditions is thus

warranted.

In future research, it would also be beneficial to

broaden the sample of observers to include

professional soldiers, reconnaissance specialists, or

personnel from other combat-oriented military

branches. This would allow for a comparative

analysis of results across different levels of

operational expertise.

A further research direction lies in evaluating the

performance of combined camouflage techniques.

Future studies could investigate the synergy between

natural materials, camouflage nets, and decoy

positions, and assess their resistance to detection by

modern sensor systems, including thermal imaging,

multispectral sensors, and unmanned aerial

reconnaissance platforms. The outcomes of such

research could be applicable not only to tactical

military scenarios but also to civilian domains, such

as infrastructure protection.

Overall, the findings confirm that the

combination of decoy positioning and natural

camouflage provides a high degree of protection,

particularly against conventional optical surveillance

systems. Although the deployment of modern sensors

such as the MOSKITO or thermal imagers enhances

the probability of detection, even these technologies

are not fully reliable when faced with well-executed

concealment techniques.

6 CONCLUSIONS

This experiment served as an initial phase in the

validation of selected camouflage techniques under

conditions approximating those of the contemporary

battlefield specifically, in open and semi-covered

terrain with seasonally limited vegetative cover. The

data obtained provided useful pilot insights that

underscore the significant contribution of natural

camouflage and decoy positions in reducing the

probability of detection, particularly with respect to

optical systems and direct visual observation. In

contrast, unconcealed positions were identified with

high accuracy, confirming the critical role of

camouflage in the protection of firing positions. For

example, the use of natural camouflage reduced

detection success rates by up to 35% across most

sensor types, while decoy positions achieved an

average concealment effectiveness of 70% based on

observer ratings.

When comparing the performance of detection

tools, thermal scope and the MOSKITO multisensory

system emerged as the most effective. Conversely,

the performance of optical devices and human

observation was strongly influenced by the observers’

individual experience and the specific conditions of

observation. Nevertheless, even advanced sensors

exhibited certain limitations when faced with well-

executed camouflage, highlighting the importance of

deliberate and adaptive concealment even against

technologically sophisticated adversaries.

The partial results also reflect the influence of

environmental and seasonal factors on camouflage

effectiveness. The experiment was conducted during

the winter months, when sparse vegetation provided

minimal natural cover. It is reasonable to assume that

denser foliage during the summer season may

significantly enhance the effectiveness of natural

camouflage techniques however, this assumption

requires further experimental verification.

Based on the results obtained so far, it can be

concluded that effective concealment in the

operational environment necessitates a combination

of multiple techniques, with an emphasis on

adaptation to current conditions. The observed

influence of the human factor including the

observers’ individual skills, knowledge, and

perceptual abilities remains a significant variable in

the detection process.

Future research should focus on expanding the

experimental framework: including a broader and

more diverse sample of observers (e.g., active-duty

military personnel), conducting trials in varied

climatic and terrain conditions, and incorporating a

wider range of detection systems, such as

multispectral sensors and unmanned aerial platforms.

The experiment was designed with a strong

emphasis on repeatability and methodological rigor.

All scenarios and conditions were meticulously

documented to enable replication by other research

teams under comparable settings. Internal validity

was supported through a controlled environment and

standardized instructions for all evaluators, while

external validity was reinforced by selecting

conditions reflective of real-world deployment

scenarios encountered by mortar units.

REFERENCES

Akdemir, E., Korkmaz, M., 2024. Camouflage as a visual

language of war: symbolic interpretation in modern

conflicts. Journal of Military Visual Culture, 12(1), s.

45–59.

Ali, M.F., Zafar, H.T., Iqbal, A., 2023. Object detection and

classification using deep learning for surveillance

applications. International Journal of Heat and

Experimental Evaluation of Camouﬂage Effectiveness Against Ground-Based Surveillance

433

Technology, 41(2), s. 625–632. https://doi.org/10.1

8280/ijht.410218

Amphenol Aerospace, 2023. The modern command and

control (C2) framework [online]. Available at:

https://www.amphenol-aerospace.com/markets/market

-connector-the-modern-command-and-control-c2-

framework

Beals, M., 2023. Artificial intelligence (AI) and machine

learning in sensor signal and image processing. Military

Aerospace Electronics [online]. Available at:

https://www.militaryaerospace.com/computers/article/

55273984/artificial-intelligence-ai-and-machine-

learning-in-sensor-signal-and-image-processing

CZ Defence, 2022. MAD21: The new camouflage pattern

for the Army of the Czech Republic will provide

effective concealment in both natural and urban

environments. [online]. Available at: https://www.czd

efence.cz

Gardony, A.L., Bryant, D.J., Healey, M.K. et al., 2022.

Target highlighting in augmented reality degrades

broader situational awareness. Journal of Cognitive

Engineering and Decision Making, 16(1), s. 34–49.

https://doi.org/10.1177/15553434211041168

Havlík, T., Blaha, M., Potužák, L., Pekař, O., Šlouf, V.

Wargaming simulator MASA SWORD for training and

education of Czech army officers. In: Proceedings of

the 16th European Conference on Games Based

Learning, ECGBL 2022. Academic Conferences

International Limited, 2022, roč. 16., s. 811-813. ISBN

978-1-914587-52-8.

Hughes, S., 2024. Artificial Intelligence in a Multidomain

Battlefield. Military Review – Journal of the U.S. Army

[online]. Available at: https://www.armyupress.a

rmy.mil/Journals/Military-Review/Online-

Exclusive/2024-OLE/Multidomain-Battlefield-AI/

Hwang, J., MA, Y., 2024. MCAM: A Military Camouflage

Dataset for Masked Target Detection. Proceedings of

the IEEE/CVF Winter Conference on Applications of

Computer Vision (WACV), s. 3125–3134.

https://doi.org/10.1109/WACV56688.2024.00308

Ivan, J., Šustr, M., Pekař, O., Potužák, L. Prospects for the

Use of Unmanned Ground Vehicles in Artillery Survey.

In: Proceedings Of The 19th International Conference

On Informatics In Control, Automation And Robotics

(ICINCO). Lisabon, Portugalsko: SCITEPRESS, 2022,

roč. 2022, s. 467-475. ISSN 2184-2809. ISBN 978-989-

758-585-2. doi:10.5220/0011300100003271

Karthiga, R., Asuntha, A., 2025. CAMOUFLAGE-Net:

Deep Camouflaged Target Detection Using YOLOv7

with Attention Mechanism. Defence Technology (In

Press).

Khoma, V., 2023. Camouflage in Ukrainian battlefield

tactics: blending, deception and adaptation. Eastern

European Security Review, 9(2), s. 103–119.

Kim, D.H., Yang, J.H., Kwon, O., 2021. Network-based

evaluation of camouflage effectiveness using clustering

of pattern similarity. Applied Sciences, 11(4), 1843.

https://doi.org/10.3390/app11041843

Li, C., Tian, Y., Zhou, M. et al., 2024. MilDetr: A

Transformer-Based Model for Detecting Camouflaged

Military Targets. Military Sensing Symposium

Proceedings. https://doi.org/10.48550/arXiv.2401.045

Li, Y., Zhang, J., Liu, H. et al., 2022. CSsub: A perceptual

model of camouflage effectiveness based on color,

texture and structure similarity. Computers & Graphics,

107, s. 84–93. https://doi.org/10.1016/j.cag.2022.0

1.004

Matthews, R., Wu, H., Xu, L. et al., 2024. Bio-Inspired

Camouflage in Autonomous Swarms and Bionic

Systems. Robotics and Autonomous Systems, 168,

104632. https://doi.org/10.1016/j.robot.2024.104632

Mulla, M.Y., Patil, M.A., Shinde, R.S., 2024. Comparative

Analysis of YOLOv5 Architectures for Detection of

Camouflaged Human Targets. Defence Science

Journal, 74(1), s. 58–67. https://doi.org/10.1442 9/dsj.

74.18654

Sedláček, M., Dohnal, F., Ivan, J., Šustr, M. Possible

approaches to assessing terrain mobility after the effects

of artillery munition. Cogent Social Sciences, 2024,

10(1), 2368096. ISSN 2331-1886. doi:10.1080/

23311886.2024.2368096

Su, L., Yu, H., Zhao, W., 2023. Advances in infrared

camouflage textiles: Flexible and adaptive materials for

stealth applications. Materials Today, 64, s. 92–105.

https://doi.org/10.1016/j.mattod.2023.05.001

Šlouf, V., Blaha, M., Müllner, V., Brizgalová, L., Pekař, O.

An Alternative Model for Determining The Rational

Amount of Funds Allocated to Defence of The Czech

Republic in Conditions of Expected Risk. Obrana a

strategie, 2023, 2023(1), 149-172. ISSN 1214-6463.

doi:10.3849/1802-7199.23.2023.01.149-172

Wei, Y., Deng, H., Zhou, Y. et al., 2021. DeepCam:

Automatic camouflage pattern generation via

convolutional neural networks and eye-tracking

validation. Computers & Graphics, 97, s. 1–12.

https://doi.org/10.1016/j.cag.2021.03.004

ZHANG, Weiguo, Zhen CHEN A Xiaowei WANG, 2021.

Review of camouflage and concealment strategies in

modern warfare. Defence Technology, 17(1), s. 90–

101. https://doi.org/10.1016/j.dt.2020.04.001

Zeng, X., Luo, Q., Chen, R., 2024. UAV-based

camouflaged target detection under low-angle

observation using hybrid CNN-Transformer networks.

Sensors, 24(3), 512. https://doi.org/10.3390/

s24030512.

ICINCO 2025 - 22nd International Conference on Informatics in Control, Automation and Robotics

434