@article{CamargoSchirrmannLandwehretal.2021,
  author    = {Camargo, Tibor de and Schirrmann, Michael and Landwehr, Niels and Dammer, Karl-Heinz and Pflanz, Michael},
  title     = {Optimized deep learning model as a basis for fast UAV mapping of weed species in winter wheat crops},
  series = {Remote sensing / Molecular Diversity Preservation International (MDPI)},
  volume    = {13},
  journal   = {Remote sensing / Molecular Diversity Preservation International (MDPI)},
  number    = {9},
  publisher = {MDPI},
  address   = {Basel},
  issn      = {2072-4292},
  doi       = {10.3390/rs13091704},
  pages     = {19},
  year      = {2021},
  abstract  = {Weed maps should be available quickly, reliably, and with high detail to be useful for site-specific management in crop protection and to promote more sustainable agriculture by reducing pesticide use. Here, the optimization of a deep residual convolutional neural network (ResNet-18) for the classification of weed and crop plants in UAV imagery is proposed. The target was to reach sufficient performance on an embedded system by maintaining the same features of the ResNet-18 model as a basis for fast UAV mapping. This would enable online recognition and subsequent mapping of weeds during UAV flying operation. Optimization was achieved mainly by avoiding redundant computations that arise when a classification model is applied on overlapping tiles in a larger input image. The model was trained and tested with imagery obtained from a UAV flight campaign at low altitude over a winter wheat field, and classification was performed on species level with the weed species Matricaria chamomilla L., Papaver rhoeas L., Veronica hederifolia L., and Viola arvensis ssp. arvensis observed in that field. The ResNet-18 model with the optimized image-level prediction pipeline reached a performance of 2.2 frames per second with an NVIDIA Jetson AGX Xavier on the full resolution UAV image, which would amount to about 1.78 ha h(-1) area output for continuous field mapping. The overall accuracy for determining crop, soil, and weed species was 94\%. There were some limitations in the detection of species unknown to the model. When shifting from 16-bit to 32-bit model precision, no improvement in classification accuracy was observed, but a strong decline in speed performance, especially when a higher number of filters was used in the ResNet-18 model. Future work should be directed towards the integration of the mapping process on UAV platforms, guiding UAVs autonomously for mapping purpose, and ensuring the transferability of the models to other crop fields.},
  language  = {en}
}
@article{ZudePflanzSpinellietal.2011,
  author    = {Zude, Manuela and Pflanz, Michael and Spinelli, Lorenzo and Dosche, Carsten and Torricelli, Alessandro},
  title     = {Non-destructive analysis of anthocyanins in cherries by means of Lambert-Beer and multivariate regression based on spectroscopy and scatter correction using time-resolved analysis},
  series = {Journal of food engineering},
  volume    = {103},
  journal   = {Journal of food engineering},
  number    = {1},
  publisher = {Elsevier},
  address   = {Oxford},
  issn      = {0260-8774},
  doi       = {10.1016/j.jfoodeng.2010.09.021},
  pages     = {68 -- 75},
  year      = {2011},
  abstract  = {In high-value sweet cherry (Prunus avium), the red coloration - determined by the anthocyanins content - is correlated with the fruit ripeness stage and market value. Non-destructive spectroscopy has been introduced in practice and may be utilized as a tool to assess the fruit pigments in the supply chain processes. From the fruit spectrum in the visible (Vis) wavelength range, the pigment contents are analyzed separately at their specific absorbance wavelengths. A drawback of the method is the need for re-calibration due to varying optical properties of the fruit tissue. In order to correct for the scattering differences, most often the spectral intensity in the visible spectrum is normalized by wavelengths in the near infrared (NIR) range, or pre-processing methods are applied in multivariate calibrations. In the present study, the influence of the fruit scattering properties on the Vis/NIR fruit spectrum were corrected by the effective pathlength in the fruit tissue obtained from time-resolved readings of the distribution of time-of-flight (DTOF). Pigment analysis was carried out according to Lambert-Beer law, considering fruit spectral intensities, effective pathlength, and refractive index. Results were compared to commonly applied linear color and multivariate partial least squares (PLS) regression analysis. The approaches were validated on fruits at different ripeness stages, providing variation in the scattering coefficient and refractive index exceeding the calibration sample set. In the validation, the measuring uncertainty of non-destructively analyzing fruits with Vis/NIR spectra by means of PLS or Lambert-Beer in comparison with combined application of Vis/NIR spectroscopy and DTOF measurements showed a dramatic bias reduction as well as enhanced coefficients of determination when using both, the spectral intensities and apparent information on the scattering influence by means of DTOF readings. Corrections for the refractive index did not render improved results.},
  language  = {en}
}