In this paper, the problem of process automation of wheat seeds germination evaluation based on bioelectric signals is investigated. A technique is proposed, and software for obtaining the database of grain germination samples for possible use in further experiments is presented. Experimental data processing starts with lowfrequency filtration performed by a moving average algorithm with further exclusion of all the rejected data. The processing is followed by reduction of the rest data to point with the lowest bioelectric impulse and calculation of parameters mean values and confidence intervals. Evaluation of grain samples with unknown germination is performed with the following parameters: the amplitude of the pulse leading edge and its length, the amplitude of the pulse trailing edge and its length, the amplitude of the pulse initial value, and the amplitude mean value at the phase of repolarization. Comparison of the given parameters and production of resulting indices allows the grain germination to be evaluated. The proposed technique and software allows the database of grain germination samples to be obtained and semi-automatic samples identification to be performed.