Variable Fertilizer Recommendation for Grass Production by Image-based Growth Status

Md Shaha Nur Kabir, Sun Ok Chung, Bo Eun Jang, Yong Joo Kim, Geung Joo Lee, Kyeong Hwan Lee, Takashi Okayasu, Eiji Inoue

Research output: Contribution to journalArticlepeer-review

3 Citations (Scopus)


Mowing and fertilizer application are important field operations for proper grass growth management. Because grass growth is not uniform across fields, variable rate fertilization may be an effective method for achieving high grass yields and enhancing fertilizer use efficiency. The objective of this research was to recommend variable rate fertilization for sod production fields depending on site-specific grass growth levels estimated by CCD camera images. CCD images were acquired from three different sod production fields and calibration relationships were obtained between the sensor measurements and the grass growth levels. Based on the calibrations, fertilizer application rates were site-specifically recommended based on expert experience. The grass growth levels and the fertilizer requirements were mapped for the three fields through geostatistical analyses. For the analysis of the spatial variation of growth level and mapping, geostatistical methods such as variogram modeling and Kriging interpolation were used. Linear relationships were obtained for the calibrations between the image-processed results and grass growth levels. Inverse linear relationships between fertilizer recommendations and grass growth levels were obtained with R2of 0.966, 0.971, and 0.944 for 90%, 70%, and 50% grass growth coverage fields. Exponential models were found to best fit data for 90% and 50% grass growth coverage fields with R2 of 0.843 and 0.780, respectively. Whereas, a spherical model was found to best fit data for 70% growth grass coverage field with R2 of 0.968. Turfgrass growth levels were divided into 0 (0%) to 10 (100%) levels, and accordingly, fertilization levels of 10 (100%) to 0 (0%) levels were recommended. The interpolated maps showed a recommended fertilizer application range of 0.5 (5%)~1.0 (10%), 0.5 (5%)~1.5 (15%), and 2.0 (20%)~ 4.0 (40%) levels in the middle areas for 90%, 70%, and 50% growth coverage fields, respectively. Whereas, ranges of 3.5 (35%)~4.7 (47%), 3.0 (30%)~4.0 (40%), and 2.0 (20%)~3.0 (30%) levels were recommended in the edge areas of 90%, 70%, and 50% growth coverage fields, respectively. This study could contribute to increase fertilizer use efficiency, reduce environmental contamination, and improve grass quality and growth if the recommended fertilizers are variably applied.

Original languageEnglish
Pages (from-to)145-155
Number of pages11
JournalJournal of the Faculty of Agriculture, Kyushu University
Issue number1
Publication statusPublished - 2019

All Science Journal Classification (ASJC) codes

  • Biotechnology
  • Agronomy and Crop Science


Dive into the research topics of 'Variable Fertilizer Recommendation for Grass Production by Image-based Growth Status'. Together they form a unique fingerprint.

Cite this