Effective risk perception is paramount in driving safety, particularly in assessing potential collision risks. Gaze regulation plays a vital role in providing visual feedback for such risk perception; however, understanding how human drivers adjust their gaze behaviors to avoid traffic hazards remains insufficiently explored. This study investigates gaze regulation for risk perception to offer valuable insights. Eleven participants engaged in a simulated collision avoidance task, decelerating at rates of − 2.5, − 5.0, or − 7.5 m/s2 from initial velocities of 40, 60, or 80 km/h to avoid collisions with a parked vehicle. Gaze regulation was analyzed using a novel application of discrete wavelet transform to denoise the gaze signal, allowing access to the spatial distribution of gaze regulation. Results revealed a consistent downward-constrained gaze regulation when drivers decelerated to avoid collisions, suggesting a primary function of vigilant monitoring of potential collision risks. The implications of gaze regulation as a key indicator of driver braking intent are discussed, particularly in mitigating conflicts between human operators and automated driving systems. Such insights could inform the development of human-centered automated driving systems.