To realize the recognition of surface damages with a low missed detection rate on large aperture optical elements in a high power laser driver and according to the low signal-to-noise ratio characteristic of micro-size damages, an adaptive detection algorithm is proposed based on the improved local signal intensity ratio. The signal intensity difference between the damages and their neighbored non-damage area in signal images is adopted to construct a filter template for the adaptive local enhancement of signal images. Thus, the damage signal intensity can be enhanced effectively and the signal-to-noise ratio of damage on signal images can also be improved significantly. As for the seed images, the seed selection and the adaptive regional growth are performed, and the complete segmentation of damage regions is finished by the accurate extraction. The improved local area signal intensity ratio algorithm can be used to effectively recognize the small-size damage points with a low signal-to-noise ratio. Under the total internal reflection lateral illumination in dark fields, the damage points with size of about 30 μm can be recognized. The missed detection rate is lower than that of the existing local area signal strength algorithm. The results show that the missed detection rate is below 0.4% for damage points with equivalent diameter of above 50 μm.