Refine
Document Type
- Article (2)
- Part of a Book (1)
- Conference Proceeding (1)
Has Fulltext
- no (4)
Is part of the Bibliography
- yes (4)
Keywords
- HYDRUS-1D (1)
- N-analysis plants (1)
- classification (1)
- color image processing (1)
- greenhouse production (1)
- growing media (1)
- histogram analysis (1)
- peat substrate (1)
- starch detection apples (1)
- water content (1)
Institute
- Fakultät AuL (4)
Evaporation from growing media significantly contributes to increasing the humidity in greenhouses. The effects of a pine bark mulch cover on substrate evaporation was evaluated with different pot experiments. The obtained data have been tested within the water balance model HYDRUS-1D, which was originally developed for mineral soils. Objective of this study was to test the performance of HYDRUS-1D to describe evaporation in plant containers and to evaluate the effect of pine bark as cover layer or layers within growing media. Application of pine bark in combination with peat substrate reduced evaporation up to 50% depending on position, thickness of mulch layer and water content of the substrate. The highest reduction in evaporation was measured in a dry substrate which is covered with 4 cm pine bark. The HYDRUS-1D model describes evaporation from growing media in combination with layers of pine bark correctly as long as hysteresis of the water retention curve and vapor flow is considered in the model.
Computer-image processing becomes more and more important in the analysis of data in biological and agricultural research and practice. However, robust image processing is highly de pendent on the histogram analysis algorithms used and the quality of the data being processed. The algorithm presented here aims to improve the accuracy of the classification of image data generated under complex boundary situations and inconsistent lighting conditions. Using the example of the determination of nitrogen content of tomato leaves and the qualitative determination of starch con tent of apples on the basis of color image processing, we showed that the developed algorithm is able to perform a robust classification and represents an improvement to simple histogram analysis.