Schizophrenia, which affects 1% of the world’s population, needs to be detected early to prevent severe cases. This research used a non-invasive electroencephalography (EEG) technique to identify abnormalities in brain activity associated with schizophrenia, along with a combination of wavelet transform and convolutional neural network (CNN) algorithm. The dataset included 81 subjects: 49 with schizophrenia and 32 healthy individuals. To address the issue of class imbalance and mitigate potential bias towards the more prevalent class (SZ), we increased the number of samples from the underrepresented class (HC) by applying image augmentation techniques. The finding demonstrated that 90% accuracy was attained using the suggested approach. These findings show that the suggested technique is a viable strategy for identifying EEG patterns linked to schizophrenia. The use of oversampling techniques to balance the class distribution in the dataset and the mix of wavelet transform and CNN algorithm allowed the effective extraction and analysis of relevant features from the EEG signals.