The invention relates to the technical field of
laser medical cosmetology, discloses a
laser hair removal method integrating
hair follicle position identification, and aims to solve the problems of low efficiency, non-uniform coverage and
safety risk caused by difficulty in accurately identifying three-dimensional distribution of hair follicles, incapability of dynamically adjusting
laser parameters and lack of closed-loop feedback of existing
hair removal equipment. According to the method, near-
infrared optical coherence tomography and high-resolution color images are synchronously collected, after
noise reduction and registration, a deep
convolutional neural network is input to segment a
hair follicle area, and a three-dimensional
space model of the
hair follicle is reconstructed based on depth information; meanwhile, the
skin temperature is monitored in real time through a thermal imager, and laser output is adjusted in a closed-loop mode in combination with a proportional-integral-differential controller, so that the
skin temperature is always lower than 45 DEG C;
after treatment, the recognition model is incrementally updated through effect
evaluation data, and the subsequent treatment precision is improved. According to the technical scheme, precise targeting and personalized
energy delivery of the hair follicles are achieved, and the effectiveness and safety of
hair removal are remarkably improved.