Back to Search
Start Over
Rapid detection of total and ammonium nitrogen in pit mud by hyperspectral imaging combined with PSO-LSSVM.
- Source :
-
Infrared Physics & Technology . Jun2024, Vol. 139, pN.PAG-N.PAG. 1p. - Publication Year :
- 2024
-
Abstract
- • The content of total nitrogen and ammonia nitrogen in pit mud was rapidly detected. • The characteristic wavelength is extracted by the combination of VCPA-IVSO. • Four prediction models (ELM, SVM, LSSVM, PSO-LSSVM) are established. Total nitrogen and ammonium nitrogen play a pivotal role in preserving the soil's nutrient ecosystem, and rapid determination of their levels in the soil can help improve soil utilization. In this study, hyperspectral imaging technology(HSI) was used to achieve rapid determination of total and ammonia nitrogen in pit mud. Both types of spectral data were pre-processed and the characteristic wavelengths were screened using a combination of variable combined population analysisand iterative variable subset optimization (VCPA-IVSO). Four prediction models (ELM, SVM, LSSVM, PSO-LSSVM) based on characteristic wavelengths and full wavelengths were developed. The results showed that the PSO-LSSVM model with the NIR characteristic wavelengths was more effective in predicting total and ammonia nitrogen in pit sludge (total nitrogen: R c 2 = 0.9999, RMSEC = 0.0014, R p 2 = 0.9892, RMSEP = 0.0111, RPD = 9.6225; ammonia nitrogen: R c 2 = 0.9941, RMSEC = 4.3004, R p 2 = 0.9644, RMSEP = 9.8962, RPD = 5.3).The distribution of total nitrogen and ammonia nitrogen in the region of interest is visually analyzed using the determined best model. These results underscore the potential of HSI as a rapid and non-destructive method for detecting total and ammonia nitrogen levels in pit mud. [ABSTRACT FROM AUTHOR]
- Subjects :
- *HYPERSPECTRAL imaging systems
*MUD
*SPECTRAL imaging
*NITROGEN
*AMMONIUM
Subjects
Details
- Language :
- English
- ISSN :
- 13504495
- Volume :
- 139
- Database :
- Academic Search Index
- Journal :
- Infrared Physics & Technology
- Publication Type :
- Academic Journal
- Accession number :
- 177453474
- Full Text :
- https://doi.org/10.1016/j.infrared.2024.105298