System based on RGB sensor for detecting camouflage in military environments

The subject of the article is the development of an inexpensive optical system based on an RGB sensor for de-tecting camouflaged objects in a military environment. The aim of the research is to create a prototype hardware and software system for detecting camouflage based on an RGB sensor operating in the visible light spectrum, with subsequent experimental verification of its effectiveness in various natural conditions. The system is characterized by low power consumption, mobility, and can be integrated into portable or un-manned reconnaissance platforms. To achieve the set goal, the following tasks were solved: development of the hardware structure of the sensor system based on available components; implementation of algorithms for background modeling, color anomaly detection, and noise filtering using colorimetric analysis in RGB space; conducting a series of field experiments in various natural conditions – wooded areas, steppe zones, and rocky landscapes; evaluating the accuracy, stability, and limitations of the proposed solution in comparison with thermal imagers and infrared cameras with artificial intelligence. The methodology is based on the use of a TCS34725 RGB sensor with a built-in infrared filter and a 16-bit ADC in combination with an ESP32 microcontroller, which provides real-time data processing, wireless data transmission, and autonomous operation. The detection algorithm is based on the formation of a color background profile, analysis of deviations by Euclidean distance, normalization of RGB values, median filtering, and adaptive threshold determination, which ensures resistance to environmental changes. Test results showed that the proposed system is capable of accurately identifying digital «pixel» camouflage in wooded areas (86 %), multicam camouflage in steppe conditions (78 %), and plain olive camouflage on rocky backgrounds (65 %). A comparison with thermal imaging and IR systems confirmed the significant advantages of the RGB solution in terms of cost (less than $ 20), energy efficiency, and mobility, although it can only operate during daylight hours. The conclusions emphasize the feasibility of using RGB detection systems as a cost-effective auxiliary tool for reconnaissance and security tasks. The scientific novelty of the work lies in the integration of a simple and affordable RGB sensor with adaptive algorithms for background modeling and color deviation analysis, which allows the creation of low power platforms for camouflage detection. Unlike traditional approaches, the proposed system opens up prospects for deploying a distributed network of inexpensive sensor nodes or integrating them into unmanned aerial vehicles for operational monitoring of large areas.