Acoustic detection of boiling water stages
Autor: | Milos Bjelic, Dragana Sumarac Pavlovic, Lara Kasca |
---|---|
Rok vydání: | 2020 |
Předmět: |
Computer science
business.industry Gaussian Pattern recognition Support vector machine Statistical classification symbols.namesake ComputingMethodologies_PATTERNRECOGNITION Boiling Classifier (linguistics) Pattern recognition (psychology) Range (statistics) symbols Mel-frequency cepstrum Artificial intelligence business |
Zdroj: | 2020 28th Telecommunications Forum (TELFOR). |
Popis: | This paper aims to identify the phases of the cooking process, specifically, in this case, the boiling water stages, using the knowledge of pattern recognition theory and mel-frequency cepstral coefficients. For the needs of the work, a database of boiling water images was created, which, after manual marking was used in training several classification algorithms, the best results of which were achieved by a medium Gaussian SVM with a high degree of accuracy after optimization. The classifier/model designed in this way, with possible small changes, can be applied in practice, that is, realized and installed in the stoves of future generations. In addition to practical applications, the paper indicates the fusion potential of the areas mentioned here as well as a wider range of applications of mel coefficients than the standard one. |
Databáze: | OpenAIRE |
Externí odkaz: |