lkj & m".kdfVca/kh; pØokr dh rhozrk dk vkdyu djus ds fy, lewps fo’o esa O;kogkfjd :i ls mi;ksx dh tkus okyh M~oksjd rduhd esa mixzg ls izkIr fp=ksa dk mi;ksx fd;k tkrk gSA blesa O;ofLFkr laogu ds laca/k esa fo’ys"kd }kjk fd, x, foospu lfgr dqN izk;ksfxd ekunaMksa ds vk/kkj ij mixzg ls izkIr fp= ds iSVuZ dh igpku dh tkrh gSA fofHkUu fo’ys"k.k dsUnzksa }kjk fdlh ,d pØokr dk vkdyu djus esa gksus okyh fo"k;ijd foospu laca/kh folaxfr;k¡ daI;wVj ij vk/kkfjr ,yxksfjFe ds ek/;e ls de gqbZA bl la’kksf/kr rduhd dks fodflr fo"k;ijd M~oksjd rduhd ¼,- vks- Mh- Vh-½ dgk x;k vkSj ;g iw.kZ fodflr m".k dfVca/kh; pØokrksa ds fy, mi;qDr gSA bl 'kks/k&Ik= esa o"kZ 2004 esa vk, rhu m".kdfVca/kh; pØokrksa ds laca/k esa ,- vks- Mh- Vh- ds dk;Z & fu"iknu dk ewY;kdau fd;k x;k gSA buds rqYukRed fo’ys"k.k ls ;g irk pyk fd ,- vks- Mh- Vh- rduhd M~oksjd rduhd ds vk/kkj ij fd, x, pØokr dh rhozrk ds vkdyuksa] tks m".kdfVca/kh; fo’ys"k.k dsUnzksa ds mixzg ls izkIr fp=ksa ds fo’ys"kdksa }kjk O;kogkfjd :Ik ls rS;kj fd, x,] ds eqdkcys dh jghA
Dvorak technique operationally used all over the world for estimating the tropical cyclone intensity is based on satellite observations. It involves image pattern recognition based on certain empirical rules along with the analyst interpretation of organized convection. The computer-based algorithm can minimize these subjective judgement discrepancies between different analysis centers estimating the same storm. This modified version is called Advanced Objective Dvorak Technique (AODT) and which is applicable for well-developed tropical cyclones. In this paper the performance of the AODT is evaluated on three cases of the year 2004 tropical cyclones. Comparative analysis indicates the technique to be competitive with, the Dvorak-based intensity estimates produced operationally by satellite analysts from tropical analysis centers.
Comparative analysis of selected facial recognition algorithms
