In this study, COLORIMETRIC CALIBRATION of scanner has been done via perceptron neural network with three or jour layers by back propagation algorithm for colored polyester fabrics. The results obtained for random training samples are not satisfactory but application of selective training samples for L*a*b* or RGB leads to good results, with better results obtained for the L* a* b* method On the other hand, the color differences between calculation XYZ and real XYZ for unknown samples, are not only in agreement with the results of polynomials and regression methods, but are also better than the results obtained in previous studies where neural network had been used for COLORIMETRIC CALIBRATION of scanner.`

Using 48016 synthetic maximum Wood-Anderson amplitudes read from waveforms of 2650 events recorded by stations of Iranian Seismological Center (IRSC, irsc. ut. ac. ir), Iranian National Seismograph Network (INSN, www. iiees. ac. ir) and temporary seismic networks belong to Institute for Advanced Studies in Basic Sciences (IASBS, iasbs. ac. ir), the empirical attenuation curve ( 0  log A ) for local magnitude of Iran has been calculated as follows: 0 log (1. 556 0. 06) log (0. 001637 0. 0009) ( 100) 3 100 R A R              where R is hypocentral distance in km and 0 A is maximum displacement amplitude of shear wave in millimeter. The empirical attenuation relationship is valid for hypocentral distances equal or smaller than 800 km. ML amplitude is the maximum amplitude observed on a Wood-Anderson (W-A) seismogram. We manually picked the maximum amplitudes on the shear window of synthetic W-A seismograms having S/N of larger than 5. We calculated synthetic W-A seismograms by removing the instrument response of each record and convolving the resulting signal with the response of the standard W-A torsion seismograph. We assumed a static magnification of 2080 for the W-A instrument. The selected ML amplitudes are belonging to events at hypocentral distance of 10 to 800 km. Except for the Makran and South Caspian Basin regions, the ray coverage of the ML amplitude covers properly the whole Iranian Plateau. To reduce the scatter of magnitude residues and ensure a reliable calculation of the attenuation curve, the selected events belong to 45 precisely relocated seismic clusters with location uncertainties of 5 km or less. The cluster approach produces redundancy in amplitudes arriving from a cluster to a given station. The redundancy will facilitate easy recognition and removal of possible outliers and thus provide a reliable estimate for the magnitude station correction. The magnitude station corrections attempts to absorb the regional attenuation difference relative to that dictated by average attenuation relationship derived in this work. The calculated attenuation curve shows a larger geometrical spreading for hypocentral distances closer than 100 km, representing a super-spherical geometrical spreading, and a smaller value for intrinsic attenuation for distances farther than 200 km once compared with the currently used ML relationship of Hutton and Boore (1987). Excluding amplitudes with hypocentral distances smaller than 60 km results in a geometrical spreading coefficient close to spherical spreading, emphasizing the importance of near distances data on accurate estimation of the geometrical spreading value. The difference in the attenuation parameters between our results and those of Hutton and Boore (1987) relationship clearly indicates the crustal disparity of Iranian Plateau and southern California. This necessitates using the new attenuation relationship for Iran. We calculated the local magnitude empirical attenuation relationship by inverting the amplitude data set for the geometrical spreadin]g and intrinsic attenuation. We did not consider magnitude station corrections in our inversion to avoid any tradeoff between the station corrections and attenuation parameters. We have shown that the magnitude residuals calculated by our local magnitude empirical relationship do not vary systemically versus hypocentral distance or magnitude. Due to the cluster-wise approach in selection of our events and partially because of the precise location of the selected events, the standard deviation of magnitude residues is about 0. 19, significantly smaller than those reported by others. We calculated the station corrections by averaging the magnitude residual in each station. The station corrections vary between-0. 44 to 0. 32. Generally, stations located in Zagros, Alborz and north west of Iran have negative station correction representing amplitude amplification in these regions relative to central Iran and north east of Iran. The new attenuation relationship provides better estimates for the attenuation parameters and especially provides precise magnitudes at close hypocentral distances. By time, the expansion of Iranian seismic networks reduces the average distance spacing of Iranian seismic stations and thus usage of better local magnitude formula such as ours becomes more important.