Lick indices are an important metrics for measuring spectral line strength. In the current methods for calculating the Lick index, the wavelength range of the spectral line profile is determined by taking the mean over a fixed interval. This approach often fails to accurately capture the actual strength of individual spectral lines, thereby compromising the accuracy and reliability of the Lick index. To address this problem, this paper introduces an adaptive spectral line measurement method based on local trend characteristics. Firstly, a boundary factor n is defined according to the spectral line characteristics to limit the extreme range of the spectral line profile. Secondly, the slopes on blue and red bands of the core wavelength of the spectral line are calculated to capture the trend information. According to the increasing or decreasing trends changes, n flux peaks (or valleys) closest to the core wavelength on both sides are obtained respectively. Subsequently, the maximum (minimum) flux values and the corresponding wavelengths are selected as the boundary points of the spectral line profile. Finally, based on these boundary points, the line index is calculated through either Equivalent Width (EW) or Magnitude (Mag) formulations, and this value is used to measure the strength of the spectral line feature. By measuring the strength of Ca4227, Hβ, Mgb, and H alpha absorption lines in F, G, and K-type stellar spectra, this paper compares and analyzes the differences between the adaptive and fixed interval measurement methods from three perspectives. First, the scatter distribution map of the Adaptive EW values obtained by the adaptive method and the EW values obtained by the fixed interval method shows that most data points exhibit certain aggregation trends, but are not completely distributed along the diagonal, indicating that both methods are stable. Still, there are systematic differences in the calculation results. Second, the statistical results of two methods show that the mean and standard deviation of Adaptive EW are higher than those of EW, indicating that the adaptive method is more sensitive in capturing spectral detail changes, can capture spectral line changes caused by factors such as stellar individual differences and observational condition fluctuations, and thus more truly reflects the strength of the spectral line. Finally, by visually inspecting the spectral line profile range obtained by the two methods, it is shown that the adaptive method can dynamically adjust the position of the boundary points according to the actual situation of the spectral line, thereby more accurately determining the profile range of the spectral line. Therefore, the adaptive profile boundary detection method establishes an effective approach for quantifying line strengths in individual spectra.