The gamma-ray total absorption facility (GTAF) composed of 40 BaF2 detection units is designed to measure the cross section data of neutron radiation capture reaction online, in order to comply with the experimental nuclear data sheet. Since 2019, several daunting experiment results have been analyzed and published, and we have found that one of the most important sources of experimental background is the initial α particles emitted by the BaF2 crystal, which is the core component of GTAF detection unit .Considering the current industrial manufacturing process capabilities, the impurities of Ra and its compounds cannot be completely removed from the BaF2. Developing data analysis algorithms to eliminate the influence of alpha particles in experimental data has become a key aspect. In this work, in order to meet the needs of data acquisition, online measurement and analysis of neutron radiation cross section, the GTAF data acquisition system adopts a full waveform acquisition method, which results in a large number of data recorded, transmitted, and stored during experiment, which also affects the uncertainty of the cross-section data. The number of data stored in the online experiment is about 118 MB/s, resulting in a long dead time.Based on the signal waveform characteristics of the BaF2 detection unit, in order to solve the aforementioned problems, three methods, namely the ratio of fast component to total component, pulse width, and time decay constant, are used to identify and distinguish α particles and γ rays. The quality factor FOM is utilized as an evaluation value and several experiments are conducted using three radioactive sources (22Na, 137C, 60Co) for verification.Due to the slow components of BaF2 light decay time being about 620 ns, the waveform pulse should essentially return to baseline at approximately 1900 ns to 2000 ns, allowing for the complete waveform of the γ rays signal to be captured at that moment, which may provide the best energy resolution. Therefore, in the online experiment, the integration length for the energy spectrum is chosen to be 2000 ns in this work.The quality factor is 1.19–1.41 from the fast total component ratio (fast component 5 ns, total component 200 ns) method, 0.94–1.04 from the pulse width (10% peak) method, and 0.93–1.07 from the time attenuation constant method. Through the quantitative analysis of quality factor and the comparison of energy spectrum, it is determined that the fast total component ratio method has the best effect and can effectively remove the background of α particles.The next step is to upgrade the online experimental data acquisition system to reduce the quantity of experimental data and the uncertainty of cross section data. The experimental data that need to be recorded should be the crossing threshold time (for the time-of-flight method) and the amplitude integration values of 5 ns after the threshold (for the fast component), 200 ns after the threshold (for the total component), and 2000 ns (for the energy) for each signal waveform, as well as the number of related detection units. The above information should be sufficient to complete online processing of experimental data, including the processing of the α particle background and (n,γ) reaction data. It is estimated that the data acquisition rate of the upgraded system will decrease from 118 MB/s to 24 MB/s, which can significantly reduce the dead time of the data acquisition system, thereby improving the accuracy of cross section data.