Humans have had the ability to recognise items for hundreds of years, possibly even since they first appeared on Earth. Humans use their senses of sight, smell, hearing, taste, and touch to determine the identities of things in their immediate surroundings. Transmission of sensory organ impulses to the brain allows for the latter's processing and interpretation of the data. That's why repetition is so important for learning. The learned knowledge is put to use in a wide variety of ways, from the mundane to the crucial, such as in the areas of security, surveillance, traffic monitoring, etc. Due to the short range of the human senses and the potential dangers of working in some environments, this approach has spatial and temporal constraints. Seismic signal processing employs the wavelet transform for detection and classification. However, picking the mother wavelet is tricky because it depends on how well it fits the original signal. As a result, the empirical wavelet transform (EWT) is investigated in this research paper as a potential adaptive wavelet transform for detection and classification. In the same way as EWT does, the mother wavelet is chosen mechanically with reference to the input signal. This study shows that EWT-based detection and classification methods perform better than STFT-based methods.