Guido, Rodrigo Capobianco [UNESP]Pedroso, Fernando [UNESP]Contreras, Rodrigo Colnago [UNESP]Rodrigues, Luciene Cavalcanti [UNESP]Guariglia, Emanuel [UNESP]Neto, Jogi Suda [UNESP]2022-04-292022-04-292021-10-01Digital Signal Processing: A Review Journal, v. 117.1051-2004http://hdl.handle.net/11449/231475This article introduces the Discrete Path Transform (DPT). Designed to serve as a new tool for handcraft feature extraction (FE), it improves the elementary analysis provided by signal energy (E) and enhances the humble spectral investigation granted by zero-crossing rates (ZCRs). C/C++ source-codes to realize both the DPT direct and inverse (IDPT) forms are presented together with a few hypothetical numerical examples and an application involving general signal analysis, artefact removal from biomedical signals, and spoken word recognition (SWR), thus demonstrating how useful and effective the proposed transform is. Brief comparisons with Teager Energy Operator (TEO) and a list of important references were also included.engArtefact removalFeature extraction (FE)Pattern recognition (PR)Signal energy (e)Spoken word recognition (SWR)Zero-crossing rate (ZCR)Artigo10.1016/j.dsp.2021.1031582-s2.0-85109965962