Identification of Mineral Deficiency in Rice Crop based on SVM in Approach of K-Means & Fuzzy C-Means Clustering

Author Name(s): *Prabira Kumar Sethy, Chhavi Kumari, NaliniKanta Barpanda, Baishalee Negi, Santi Kumari Behera, Amiya Kumar Rath
Author Email: prabirsethy.05@gmail.com

Abstract

To setup a quick, error free and feasible method for diagnosing rice crop nutrition is very necessary for nutritional management. In this paper the support vector machine in approach of K-Means and Fuzzy C-Means clustering is used to identify mineral deficient leaf of Rice Crop. The images are collected from the database of International Rice Research Institute (IRRI). From these images 13 features were extracted using RGB mean value function and Regionprops. These function are used to identify different mineral deficient samples such as Nitrogen, Potassium, Phosphorus, Magnesium and Zinc. The combination of SVM classifier with K-Means and Fuzzy C-Means clustering, the mineral deficient leaf image of rice crop was identified with accuracy of 85% and 92% respectively. In context of computational time the Fuzzy C-Means Clustering with SVM achieve better performance than the K-Means clustering with SVM. This proposed technique plays a significant role to identify mineral deficiency in Rice Field by implementing in electronic gadgets such as Smartphone, Tablet etc.

INTRODUCTION

Rice shows obvious symptoms when suffering from Nitrogen (N), Potassium (K), Magnesium (Mg), Phosphorous (P) and Zinc (Zn) nutritional deficiencies, and these symptoms are the basis of rapid morphological diagnoses in the field. Morphological diagnoses require large amounts of experience. This method cannot be quantified and suffers from poor operability; farmers find it very difficult to expertly use this method. Diagnostic methods extract information from the nutritional stress rice effectively and significantly, which can be used to automatically identify the nutrition status of rice [1]. Mineral deficiency mostly appears on the leaves of rice crop including off-colored leaves, abnormal shape of leaves and also breaks down of a certain part of leaves. Mineral deficiency generally appears in the older leaves when nitrogen, phosphorus, potassium and magnesium are limiting [2]. Younger leaves show a deficiency when sulphur, iron, calcium, iron, calcium, zinc, copper, boron and manganese are limiting [2].  Under N deficiency, old leaves and sometimes all leaves become light green and chlorotic at the tip. Except for young leaves, which are greener, deficient leaves are narrow, short, erect, and lemon yellowish [3][4]. Under P deficiency, leaves are taper, short and stiff. Under K deficiency, dark-green plants with yellowish-brown leaf margins or dark brown ecrotic spots first appear on the tips of older leaves. Under severe K deficiency, leaf tips are yellowish brown. Older leaves change from yellow to brown[3][4].Therefore, the color and shape of the leaf and sheath can indicate the plant nutrient and health status, which is closely related to the nutrition content. Under Mg deficiency, orange-yellow interveinal chlorosis on older leaves. Green coloring appears as a “string of beads” in which green and yellow stripes run parallel to the leaf [5]. Under Zn deficiency, dusty brown spot appears on the upper part of the leaf. Leaves lose turgor and turn brown as brown blotches and streaks appear on lower leaves, enlarge, and coalesce and also sometimes white line appears along the leaf midrib and leaf blade size is reduced. The mineral deficient rice leaf images are shown in Figure 1.

In this research, scanned images of rice leaves and sheaths under mineral deficiencies and normal nutrition levels were compared, and the differences in the rice leaf and sheath characteristics under the different nutrition conditions were analyzed. K-Means clustering and Fuzzy C-Means Clustering are used for extracting the features of sample images and by use of these features; SVM identifies the different type of mineral deficient leaf image.

CONCLUSION & FUTURE SCOPE

The input images are correctly segmented and its feature vector is correctly extracted. FuzzyC-Means clustering is more accurate &noise free technique. Fuzzy C-Means with SVM is used and the deficiencies are successfully identified giving more accuracy in comparison to K-means. Here, the average accuracy of the classifier by using K-means and Fuzzy C-Means is found out to be 85% and 92% respectively. Also, the computational time in case of SVM with Fuzzy C-Means is less compared to its counterpart i.e. K-Means clustering. For more extensive analysis a large data set is required for training and testing purposes. The further research may be carry to the toxicity and also determine the degree of severity of mineral deficiency and toxicity in Rice Crop.

 

References

  1. Rafael F. Muñoz-Huerta et al. A Review of Methods for Sensing the Nitrogen Status in Plants: Advantages, Disadvantages and Recent Advances, Sensors 2013, 13, doi:10.3390/s130810823, ISSN 1424-8220.
  2. Crop nutrient deficiency and toxicities, published by MU extensions, Mussouri-Colombia.
  3. Armstrong DL (2002) Nutrient Deficiency Symptoms in Rice. Better Crops International 16, Special Supplement: 23–25.
  4. Fundamentals of Rice crop Science, published by International Rice Research Institute, Philippines, 1981.
  5. Rice Doctor, International Rice Research Institute,Philippines, www.irri.org, 2003
  6. AshishMiyatra, SheetalSolanki (2009), Disease and Nutrient Deficiency Detection in Cotton Plant, International Journal of Engineering Development and Research, Volume 2, Issue 2, ISSN: 2321-9939.
  7. Marian Wiwart et al. Early diagnostics of macronutrient deficiencies in three legume species by color image analysis, computers and electronics in agriculture 6 5 (2 0 0 9).
  8. Nisal , Chandan J. Bharambe, VidyaN.More(2011) , Detection and Analysis of Deficiencies in Groundnut Plant using Geometric Moments,International Journal of Biological, Biomolecular, Agricultural, Food and Biotechnological Engineering Vol:5, No:10.
  9. Muhammad Asraf, M. T. Nooritawati, M.S.B. Shah Rizam(2012).A Comparative Study in Kernel-Based Support Vector Machine of OilPalm Leaves Nutrient Disease. International Symposium on Robotics and Intelligent Sensors 2012 (IRIS 2012).
  10. Sanyal1, U. Bhattacharya2, S. K. Parui, S. K. Bandyopadhyayand S. Patel. Color Texture Analysis of Rice Leaves to Diagnose Deficiency in the Balance of Mineral Levels Towards Improvement of Crop Productivity.10th International Conference on Information Technology.
  11. Prabira Sethy, BaishaleeNegi, NilamaniBhoi(2017). Detection of Healthy & Defected Diseased Leaf of Rice Crop Using K-Means Clustering Technique. International Journal of Computer Applications. January 2017, Volume 157- No.1, 0975-8887.
  12. SoumiGhosh, Sanjay kumarDubey, Comparative Analysis of K-Means and Fuzzy C-Means Algorithms.International Journal of Advanced Computer Science and Applications,Vol. 4, Issue.4, 2013.
  13. C.Jobin Christ &Dr.R.M.S.Parvathi(2011), Fuzzy C-Means Algorithm for Medical Image Segmentation, 978-1-4244-8679-3/11/$26.00 ©2011 IEEE.
  14. Seyyid Ahmed Medjahed (2015), “A Comparative Study of Feature Extraction Methods in Images Classification”, IJIGSP, vol.7, no.3, pp.16-23, 2015.DOI: 10.5815/ijigsp.2015.03.03.
  15. HyeranByun&Seong-Whan Lee (2002), Applications of Support Vector Machines for Pattern Recognition- A Survey, Springer, SVM 2002, LNCS 2388, pp. 213-236.

 

 

558 total views, 1 views today

Download PDF File

About the author: tej