Melexis Launches Free Personnel Detection Algorithm for FIR Arrays

December 5, 2025, Tessenderlo, Ham, Belgium — Global micro-engineering company Melexis today announced the launch of a new personnel detection algorithm specifically designed for the MLX90642 (32x24 pixel) thermal sensor array. This algorithm enables personnel detection, accurate counting, and position location, offering an efficient and privacy-conscious occupancy sensing solution for smart buildings.

Compared to vision-based camera systems, FIR (Far Infrared) thermal sensing technology operates without the need for visible light, maintaining stable detection even in complete darkness. Furthermore, it avoids capturing identifiable images, providing greater privacy protection in sensitive environments such as offices and conference rooms. The algorithm is specifically optimized for ceiling-mounted sensors in confined spaces like elevator cabins.

The algorithm is provided as a precompiled library, supports various commonly used microcontrollers (MCUs), and features a hardware-decoupled design with low memory footprint. This facilitates flexible integration and deployment by engineers on different cost-optimized hardware platforms. The solution was developed based on Melexis's in-house expertise in FIR imaging and uses a logic-based optimization approach. It fully leverages the low noise and wide operating range characteristics of the MLX90642 sensor, ensuring stable operation across different environments and avoiding the retraining needs and performance fluctuations common with AI-based solutions.

To simplify evaluation, Melexis also provides a dedicated evaluation kit integrating the MLX90642 sensor with pre-installed firmware, allowing engineers to quickly validate the algorithm's out-of-the-box performance. This algorithm helps accelerate the development of functions like occupancy monitoring, energy management, and low-power operation in smart building systems.

Joris Roels, Market Manager at Melexis, stated that the algorithm combines proven sensor performance with ease of use, aiming to help engineers more easily integrate advanced infrared-based personnel detection technology into next-generation smart building systems.