CROPS DIAGNOSIS USING HURST EXPONENT VALUES IN FIELDS IMAGE ANALYSIS

One of the branches of sustainable agriculture is the precision farming which assumes an individual approach to each plant. The main problem encountered by the precision agriculture is to quickly acquire and analyze good quality data assessing the condition of the crop. One of the fastest growing monitoring techniques is the analysis of images obtained from cameras placed on UAV. The studies used the chaos tools to determine Hurst exponent values received from images collected during UAV flights over the fields. The obtained results of image analysis indicated the presence of a strong dependency between the Hurst exponent values and state of crops. Images showed crops which are in good standing have been seen as strong organize objects represented by the mean Hurst exponent values from 0.8 to 0.87. Crops in which occurred the destruction of plants on the collected images were estimated by the Hurst exponent between 0.41 and 0.49 values, which indicates the presence of the characteristics of chaotic changes in the distribution of color attributes. INTRODUCTION The proper application of the sustainable agriculture is only possible through receive of the accurate information about the crop status and soil fertility. The reliable information on the status of the crop is the key data needed to take further action in it (Christy, 2008). For large fields or fields unable to monitoring through traditional methods, one of the most popular method recognize of the state of crops is analysis of the images, had sent from satellite or UAV (Unmanned Aerial Vehicles). Optical recognized methods are one of the most promising for evaluation both state of crops and the soil conditions (Shapira et al., 2013). The principal advantage of the optical methods is the high speed of measures not possible for traditional ground methods (Zwiggelaar, 1998). It is well known that in the sustainable agriculture the time is the major parameter beside of the accuracy. Compared to the traditional soil or crops measurement methods, optical methods can reduce of the total of the estimation costs up to 80-90% for large areas (Nduwamungu et al., 2009).The main indicator have used for crops state evaluation has been the NDVI index (Normalized Difference Vegetation Index). NDVI index is described as relative difference of the reflectance values of infrared and red color wave (Soliman et al., 2013). For the better discrimination, the Infrared images are supplemented by images captured in visual light range. Interpreting of these images, the very significant for correctively recognize and estimation state of crops (McNairn et al., 2009). The find the proper method for the image discrimination is the important challenge for manufacturers of the image analysis systems. In order to discriminate elements contained in the image we can use its texture. A texture can be defined as the placement of individual color identifiers in the image space. In the image consisting of a set of ordered pixels with the parameters of the point A (x, y, z), where x and y represent the position of the point on the surface of the image and z is the parameter describing the color properties of the image (Zhao & Wang, 2016). During