Weight Estimation of Wheat by Using Image Processing Techniques K. Sabanci 1 , S. Ekinci 2 , A. M. Karahan 3 , and C. Aydin 4 1 Department of Electric and Electronics Engineering, Karamanoglu Mehmetbey University, Karaman, 70100, Turkey 2 Department of Mechanical Engineering, Selçuk University, Konya, 42003, Turkey 3 Vocational School, Batman University, Batman, 72100, Turkey 4 Department of Agricultural Machinery, Selçuk University, Konya, 42003, Turkey Email: kadirsabanci@kmu.edu.tr, {sekinci, caydin}@selcuk.edu.tr, alimucahit@batman.edu.tr Abstract—Today, image processing applications in agriculture seems to spread rapidly. Image processing in agriculture are used in many areas such as classification of products, the detection of weeds, crop yields and weight estimation. In this study weight estimation of bread wheat and durum wheat in different amounts was performed by using image processing techniques. Image processing techniques were applied by using Matlab software. The counting of wheat kernels in image and weight estimation was carried out. Success rates were determined by comparing estimated weights of wheat kernels and their actual weight. Index Terms—image processing, wheat, weight estimation, durum, bread I. INTRODUCTION Wheat is the most widely grown improved cultigen and it has an important role in nutrition, agricultural industry and commerce in the world. The wheat plant which has a very important position for rapidly increasing world population nourishment can be easily produced all over the world as it is depended on machined agriculture and high adaptation ability. In addition to being the raw material of wheat bread, wheat is used in the production of pastries and biscuits [1]. Physical properties of agricultural products such as length, thickness, width, surface area, bulk density, projection area is highly important in terms of engineering. Image processing techniques are in used for measurement of physical properties of agricultural products in recent years. In general terms, image processing means that manipulation and analysis of pictorial information [2]. Image processing techniques are used in various fields such as industrial, security, geology, medicine, agriculture. Image processing and artificial neural networks in agriculture are used for purposes classification in fruit color analysis, monitoring of root growth, measurement of leaf area and determining weeds etc. [3]-[7]. Sadrnia et al. [8] classified and analyzed the fruit shapes in long type watermelon using image processing. The results of their study indicated that length to width ratio and Manuscript received March 24, 2015; revised October 29, 2015. fruit area (2D) to background area ratio can be used to determine misshapen fruit. Zayas et al. [9] used image processing to discriminate wheat and non-wheat and between weed seeds and stones in the non-wheat part of a grain sample. They reported that physical separation of stones prior to the image analysis program may be necessary for satisfactory discrimination. Also Zayas et al. [10] took advantage of the image processing techniques for classification and determination of shape properties 17 different wheat varieties. They developed methodology for wheat classes and variety identification by combination of image analysis techniques with wheat hardness physical measurements. Shouche et al. [11] quantified for shape variation in 15 Indian wheat varieties by digital image analysis using custom-built software. They placed fifty wheat grains on the scanner increase-down position avoiding grain to grain contact, thereby circumventing extensive programming needed to separate touching objects and also avoiding the associated loss of information in the images. They stored images in *.tif format for further analysis. Then they determined lengths, width, thickness, environment and shape coefficients of wheat grains via an image processing program. Bacci et al. [12] transferred images of wheat grain to a computer and analyzed via image processing technique. In this way, they determined the percentage of injured seeds through this technique. Symons et al. [13] used the image processing technique to discriminate nonvitreous wheat and vitreous wheat. Sabancı et al. [14] distinguished wild rye seeds mixed into wheat using artificial neural networks and image processing techniques. In addition, they classed wheat and rye seed in the image information received from a webcam. In this study, weight estimation of bread wheat and durum wheat in different amounts was performed by using image processing techniques. By comparing obtained results with real weight was calculated the success percent of system. This study exemplifies image processing in agriculture. II. MATERIALS AND METHODS In this study, image of bread wheat and durum wheat kernels was taken by using a Logitech C905 webcam. Journal of Image and Graphics, Vol. 4, No. 1, June 2016 ©2016 Journal of Image and Graphics 51 doi: 10.18178/joig.4.1.51-54