The first seed material laboratory experiments using the Kirlian photography method were carried out in the late 60s in Alma-Ata, Russia (Inyushin et al., 1968). The results of examining wheat grain found that the glow of sprouted grain increased sharply compared to the glow of dry grain. Buadze et al. (1989) investigated the effects of the herbicide 2,4-D on the physiological state of 7day old maize seedlings. These researchers documented the changing characteristics of Gas Discharge Images (GDI) of sprouted grain. Maximum intensity shifts in GDI parameters were fixed in the wavelength range of 350 to 450 nm. Borisova et al. (2009) studied the effect of microwave treatment of rape, barley and wheat seed using GDV. Research has shown that the glow intensity of sprouted grain was related to germination potential. Priyatkin et al. (2006) investigated the ability of GDV to characterize wheat grain with no visible signs of injury. This grain was divided into three groups: 1) healthy, 2) grains with mild internal damage from the Fusarium spp. pathogen, and 3) grain with severe internal damage from this pathogen. It was noted that healthy grain was characterized by GDV maximum values for brightness distribution, shape factor and three- dimensional fractal characteristics in comparison to seed subjected to stress.

A more recent method for interpreting complex experimental data is the use of an Adaptive Neuro- Fuzzy Inference System (ANFIS) (Jang, 1993). Recently ANFIS has been successfully applied in the assessment of several agricultural problems (Akbarzadeh et al., 2009, 2009a; Arkhipov et al., 2008; Atsalakis & Minoudaki, 2007; Azamathulla et al., 2009; Cai et al., 2004; de Araújo & Saraiva, 2003; Krueger et al., 2010, 2011; Kurtener et al., 2005; Marce et al., 2004; Ostrovskij et al., 2014; Peschel et al., 2002; Tooy & Murase, 2007; Xie et al., 2007). The objective of the current study is to ultilize ANFIS for assessing seed using Corona Discharge Photography data.

2. Materials and Methods

2.1 Experimental description

Seed samples from European spruce (Picea abies L.) were gathered in the Leningrad region of Russia (Fig. 1). One seed was placed in each well of a 96-welled plate (Fig. 2). For studying the gas discharge glow from seeds, we used a GDV Camera system to record images for analysis via ANFIS. An example seed glow can be seen in Figure 3. Analysis of changes in discharge images (GDI) included calculations of amplitude characteristics, as well as geometry, brightness, fractal, and stochastic characteristics. These calculations were conducted using the GDV Scientific Laboratory software. After testing with the gas discharge visualization method, the seeds were germinated for 15 days and root length was measured daily.