Improved Plant Image Segmentation Method using Vegetation Indices and Automatic Thresholds
2
조회
0
다운로드
1
인용
0
참고문헌
키워드
초록
We address improved plant image segmentation based on histograms which requires using avegetation index and threshold. Image segmentation is the most important step for extractingtargets, such as vegetation, from images; this affects successful detection of plant information. Forty-two field images were acquired from a soybean field using an RGB camera. ThroughK-means clustering analysis, we built a new vegetation index [(b-a)00.5(1.4R-G)] andgenerated gray-scale images. Otsu and Triangle thresholds were used to convert contrast imagesto binary. Optimal threshold values were generally located between the Otsu and Trianglethreshold values. The combined threshold method shows 98.79% and 0.95% of mean accuracyand standard deviation, respectively, whereas the Otsu and Triangle method results show98.17±1.71% and 97.85±1.87%, respectively. These results show that the combined method hassignificant segmentation potential through one-way ANOVA. Then we compared the results withK-means clustering using two-sample t-test. The K-means method’s mean accuracy is98.18±1.79%, with no significant difference between the proposed and K-means methods. However, the proposed method’s processing time is 0.60±0.01 s, i.e., twice faster than theK-means method (1.72±0.24 s).
Abstract
We address improved plant image segmentation based on histograms which requires using avegetation index and threshold. Image segmentation is the most important step for extractingtargets, such as vegetation, from images; this affects successful detection of plant information. Forty-two field images were acquired from a soybean field using an RGB camera. ThroughK-means clustering analysis, we built a new vegetation index [(b-a)00.5(1.4R-G)] andgenerated gray-scale images. Otsu and Triangle thresholds were used to convert contrast imagesto binary. Optimal threshold values were generally located between the Otsu and Trianglethreshold values. The combined threshold method shows 98.79% and 0.95% of mean accuracyand standard deviation, respectively, whereas the Otsu and Triangle method results show98.17±1.71% and 97.85±1.87%, respectively. These results show that the combined method hassignificant segmentation potential through one-way ANOVA. Then we compared the results withK-means clustering using two-sample t-test. The K-means method’s mean accuracy is98.18±1.79%, with no significant difference between the proposed and K-means methods. However, the proposed method’s processing time is 0.60±0.01 s, i.e., twice faster than theK-means method (1.72±0.24 s).
참고문헌 (21)
- Bensch CN, "Interference of redroot pigweed(Amaranthus retroflexus), Palmer amaranth(A. palmeri), and common waterhemp(A. rudis) in soybean", Weed Science, 51(1), pp.37~43 (2003)
- Campbell JB., "Introduction to remote sensing" (1996), CRC Press
- Clement A, "Unsupervised segmentation of scenes containing vegetation (Forsythia) and soil by hierarchical analysis of bi-dimensional histograms", Pattern Recognition Letters, 24(12), pp.1951~1957 (2003)
- Predicting soybean yield loss in giant foxtail(Setaria faberi)and common lambsquarters(Chenopodium album)communities, "Conley SP", ~ (2009)
- Exclusive OR(XOR)and hardware random number generators, "Davies RB.", ~ (2002)
- Hernandez-Lopez JJ, "Detecting objects using color and depth segmentation with Kinect sensor", Procedia Technology, 3, pp.196~204 (2012)
- Kim Y, "Fuzzy logic control of a multispectral imaging sensor for in-field plant sensing", Computers and electronics in agriculture, 60(2), pp.279~288 (2008)
- MacQueen J., "Some methods for classification and analysis of multivariate observations", Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability, 1(14), pp.281~297 (1967)
- McLaren K., "XIII—The development of the CIE 1976 (L*a*b*) uniform colour space and colourdifference formula", Journal of the Society of Dyers and Colourists, 92(9), pp.338~341 (1976)
- Machine vision detection parameters for plant species identification. In Photonics East(ISAM, VVDC, IEMB), "Meyer GE", International Society for Optics and Photonics, 327~335 (1999)
- Meyer GE, "Digital camera operation and fuzzy logic classification of uniform plant, soil, and residue color images", Applied engineering in agriculture (2004)
- Meyer GE, "Verification of color vegetation indices for automated crop imaging applications", Computers and Electronics in Agriculture, 63(2), pp.282~293 (2008)
- Otsu N., "A threshold selection method from gray-level histograms", Automatica, 11(285), pp.23~27 (1975)
- Park HK, "Agricultural technology guide book"soybean"", Rural Development Administration, 138~217 (2012)
- Pointer MR., "A comparison of the CIE 1976 colour spaces", Color Research & Application, 6(2), pp.108~118 (1981)
- Roberts JJ, "Marine Geospatial Ecology Tools: An integrated framework for ecological geoprocessing with ArcGIS, Python, R, MATLAB, and C++", Environmental Modelling & Software, 25(10), pp.1197~1207 (2010)
- Tong JH, "Machine vision techniques for the evaluation of seedling quality based on leaf area", Biosystems Engineering, 115(3), pp.369~379 (2013)
- Woebbecke DM, "Shape features for identifying young weeds using image analysis", Transactions of the ASAE, 38(1), pp.271~281 (1995)
- Yao H, "An improved K-means clustering algorithm for fish image segmentation", Mathematical and Computer Modelling, 58(3), pp.790~798 (2013)
- Yusoff IA, "Automated two-dimensional K-means clustering algorithm for unsupervised image segmentation", Computers & Electrical Engineering, 39(3), pp.907~917 (2013)
- Zack GW, "Automatic measurement of sister chromatid exchange frequency", Journal of Histochemistry & Cytochemistry, 25(7), pp.741~753 (1977)