Application of Adaptive Camera Zoom Using the Kalaman Filter Algorithm for Low Light Conditions

Abstract

The proliferation of visual surveillance and mobile imaging technologies has created a critical need for camera systems that perform reliably in diverse lighting environments. A significant challenge persists in low-light conditions, where conventional cameras often produce images with poor sharpness, high noise, and unstable contrast, limiting their effectiveness for security and monitoring applications. This research aims to implement the Kalman Filter algorithm in an adaptive zoom camera system to improve image quality in low-light conditions. The main problem faced by conventional cameras is the instability of light intensity, which affects image sharpness and contrast. To that end, experiments were conducted using three different Android devices, namely Infinix Hot Play 11, Oppo Reno 6, and Oppo Reno 11, with shooting distances of 30 cm and 60 cm, respectively. Each device was tested using a Kalman Filter-based camera application and compared with actual measurements using a lux meter. The results of the study show that the Kalman Filter-based adaptive camera system is capable of providing light intensity estimates that are close to the actual values, with a deviation of less than 7%. This algorithm works predictively through a process of dynamic estimation and updating of lighting values, enabling it to simultaneously adjust camera exposure and focus settings. This results in sharper, more stable, and more realistic images even in low-light environments