An automated method for extracting characteristic valence-band spectral edges from ultraviolet photoelectron spectroscopy (UPS) is proposed by combining adaptive wavelet denoising with genetic-algorithm-based two-segment linear fitting. After gold-reference energy calibration and baseline correction, a Daubechies 8 wavelet is used to suppress high-frequency noise while preserving local spectral-slope changes. A characteristic edge interval is then automatically selected, and a continuous two-segment model is optimized using a genetic algorithm with the Huber loss function. Highly oriented pyrolytic graphite  is employed as a proof-of-concept sample because its high conductivity minimizes charging effects and its pronounced π-derived valence-band feature provides a distinct spectral edge. The proposed procedure yields an extrapolated characteristic edge energy of 1.895 eV relative to the Fermi level. This value is interpreted as a π-derived spectral-edge energy rather than a conventional semiconductor VBM. The method reduces dependence on manual fitting-region selection and enables automated UPS spectral-edge analysis.

Ultraviolet Photoelectron Spectroscopy,Valence Band Maximum,Wavelet Denoising,Genetic Algorithm,Data Fitting