Long-term monitoring of horizontal movement of the top of vertical elements, such as towers of historical monuments or chimneys, remains a challenge with current solutions. In such structures, which are usually of high historical and cultural value, ageing and structural degradation directly impact their horizontal movement. This study presents a novel solution based on computer vision that allows the monitoring of horizontal movements with very good accuracy (uncertainty less than ±0.26 mm), high measurement speed (up to 5 Hz), and reduced cost. A real-life validation was carried out on the south spire of the Saint Mary Cathedral in Burgos, Spain. Its horizontal movements were monitored for more than a year, and some relevant events were recorded, including high-range thermal variations, strong wind events, and ringing of the bells. The results showed that the proposed solution was particularly robust for long-term monitoring of slow movements, such as those produced by thermal events or those derived from structural degradation. The white noise of the sensor was small and, for long-term measurements and small reading frequencies, it could obtain field accuracies of the order of ±0.1 mm at a recording frequency of 1/10 Hz, which further improved at lower recording frequencies.
