Abstarct: Electroencephalography (EEG) is the most common method of recording brain activity. This method provides us with lots of physiological datas. The obtained information from the electroencephalogram signal has great importance in researchs. Although this signal is combined with different artifacts, the most common of which are artifacts. The eye (EOG) is the result of blxinking and eye movements, feeding line noise and environmental conditions are also among the most important artifacts available. Since the presence of these artifacts in the EEG signal affects our judgment, it is very important to remove them. The goal here is to obtain higher quality EEG signals, which can be achieved by deleting these artifacts so that none of the original EEG signal information is deleted. Over the years, there have been various methods for removing ocular artifacts. It is presented as blxinking, but there is still no complete method that completely removes these artifacts from the brain signal without any error. In this dissertation, while reviewing the methods performed during the past years, the algorithm Empirical mode decomposition (ٍ EMD) and its improved (EEMD) algorithm is used on standard data. During this process, first using the intrinsic state experience method of signal The brain is decomposed into different frequencies and parts of the signal that blxinking have been removed are removed, and finally the desired signal is reconstructed and displayed using the Canonical correlation analysis (CCA) algorithm. Also, the simulation results on this data using the proposed method, through the signal-to-noise component (SNR) with other methods are compared and its relative improvement is displayed.