A non-destructive Ammonia-based pork freshness detection method utilizing UV-DOAS and spectral reconstruction fitting...

Ammonia (NH₃) is a signature gas for pork deterioration, but the ammonia-based method has not been widely used due to the limitation of ultra-trace measurement. A novel ammonia-based detection method based on ultraviolet differential optical absorption spectroscopy (UV-DOAS) and spectral reconstruction fitting neural network is proposed for ultra-trace ammonia measurement. The spectra are differentially processed to remove interference of slow absorption and spectral scattering. Spectral reconstruction is introduced to transform indescribable features into sine function utilizing the known absorption characteristics of NH₃, and discretize interfering signals that do not meet NH₃ absorption characteristics. The spectra after inverse reconstruction are applied to a neural network to realize concentration inversion. The detection range is 28.0-28,087.6 ppb, and the mean absolute percentage error (MAPE) and R² are 1.50 % and 0.999, respectively. Experimental results at 30 °C and 10 °C demonstrate that this method can effectively extract the NH₃ signal and realize ultra-trace measurement.