5 May 2019 in Journal of Sensors
Greenhouse crop production is growing throughout the world and early pest detection is of particular importance in terms of productivity and reduction of the use of pesticides. Conventional eye observation methods are nonefficient for large crops. Computer vision and recent advances in deep learning can play an important role in increasing the reliability and productivity. This paper presents the development and comparison of two different approaches for vision based automated pest detection and identification, using learning strategies. A solution that combines computer vision and machine learning is compared against a deep learning solution. The main focus of our work is on the selection of the best approach based on pest detection and identification accuracy. The inspection is focused on the most harmful pests on greenhouse tomato and pepper crops, Bemisia tabaci and Trialeurodes vaporariorum. A dataset with a huge number of infected tomato plants images was created to generate and evaluate machine learning and deep learning models. The results showed that the deep learning technique provides a better solution because (a) it achieves the disease detection and classification in one step, (b) gets better accuracy, (c) can distinguish better between Bemisia tabaci and Trialeurodes vaporariorum, and (d) allows balancing between speed and accuracy by choosing different models.
1 February 2018 in Astronomy & Astrophysics
Astronomy & Astrophysics (A&A) is an international journal which publishes papers on all aspects of astronomy and astrophysics
7 October 2019 in Astronomy & Astrophysics
Context. Meteor spectra are commonly interpreted using data from databases and tables. Several studies have demonstrated very sophisticated calculations of elemental compositions of meteoroid bodies based on the computation of synthetic meteor spectra or on the spectral analysis of airglow plasma containing evaporated, atomized, and ionized meteoroid matter. However, considering accuracy, reliability of computations, lack of laboratory experimental data in this field, as well as the complicated physical structure of meteor plasma, such qualitative assignment or quantitative calculations are still extensively discussed in the scientific community. Even on the laboratory level, many studies have shown the high complexity of the acquisition and interpretation of the data that are recorded with techniques of emission spectroscopy that are in fashion and philosophy similar to the spectral analysis of meteor plasma, that is, detection and quantification of the elements that are ablated from complicated multicomponent matrices. Aims. The current study is focused on the application of terawatt-class laser-induced breakdown spectroscopy (TC-LIBS) of real samples of chondritic meteorites. We recorded emission spectra with high resolution and high precision that contain spectral lines that are typical for real meteoric spectra. Experimental data were compiled in a form that is convenient for the meteoric spectra interpretation and calibration. Methods. TC-LIBS was carried out by a high-power terawatt-class laser facility, the Prague Asterix Laser System (PALS). The spectra were simultaneously recorded by an echelle high-resolution spectrograph in the UV/VIS spectral ranges and by a low-resolution spectrograph that was used for real observation of meteor spectra. We also present calculated synthetic spectra based on data from the NIST atomic spectra database. Results. We assembled etalon qualitative tables of major meteoric spectral features that can be used both for the spectral wavelength calibration of low-resolution observational instruments and for the exact interpretation of meteor spectra. The data are compared with real meteor spectra.
10 March 2016 in Monatshefte für Chemie - Chemical Monthly