Image processing and Ultra-Violet and Visible reflectance spectroscopy combined with chemometrics for discrimination as well as authentication powder and extract with anti- diabetic polyherbal formulation. Jayanta Kumar Maji * , Shikha Sharma ** , V.J.Shukla *** * Ph.D scholar (Ayu.), ** M-pharm (Ayu.) *** Head of the department Pharmaceutical Chemistry, I.P.G.T & R.A, PGT-SFC, Gujarat Ayurved University, Jamnagar, Gujarat-361008, India. E-mail: jkmaji67@gmail.com. Mob.no-9537130948 Abstract: _ The proposed anti-diabetic polyherbal formulation “Diabetogen” is composed of Indian Kino (Ht.wd.), Indian Liac (Lf.), Ram’s horn (Lf.), Fenugreek (Sd.), black berry (Sd.) and heart-leaved moonseed (St.) is categorized antioxidant rich medicament and has been clinically used in the Indian subcontinent to various permutation and combination. To establish a real time identification system in favor of non –destructive ultra-violet visible (UV-VIS) spectroscopy & preliminary image processing. The multivariate chemometrics technique principal component analysis (PCA), hierarchical cluster analysis (HCA) use and allow an overall evaluation of the significant difference between groups and discriminate the polyherbal powder and extract. The authenticated individual herbal, polyherbal pulverized powder and the dry extracts were both shifted through eighty mesh. The samples were subjected to UV-Vis diffuse reflectance spectral detection at the interval of 1 nm. The macroscopic image powders of herbal plants planning for plant identification, was carried out by L*A*B color based image segmentation. Plant powders macroscopic image shown distinct L*A*B color based segmentation for identifying sample. Samples discriminated by first-order derivative preprocessed reflectance spectra on favor of various transitions marker bands. Discrimination of the two classes of remedy was also able in natural grouping by PCA and HCA technique. An analytical method which is rapid, simple and accurate for discriminating two forms of polyherbal formulation using (UV-VIS) diffuse reflectance spectroscopy combined with well-known chemometrics method was developed along with powder image processing. KeyWords: Chemometrics, Discrimination, Diabetogen, Image processing, UV-VIS reflectance Spectroscopy. 1. Introduction: Herbal medicine (HM) and their preparations have been used widely for hundreds of years all over the world. Among the characteristics of herbal medicine preparations, is that they are all presented either as single herbs or as a combination of several herbs in composite formulae. Quality control protocols for herbal drugs have constituted much concern and at present it is a real challenge, especially with the development of herbal remedy of herbal formulation. A fingerprint can be demarcated as a characteristic profile reflecting the compound chemical composition when the sample analyzed and can be achieved by spectroscopic, chromatographic or RAPD technique. Fingerprinting profile have to be highlighted by the fundamental designation of ‘integrity’ and ‘fuzziness’ (sameness and difference) so as to chemically exemplify the HM has been investigated Welsh. [1] It is recommended that with the help of the chemical fingerprints , the authentication and identification of herbal medicine can be conducted accurately, ‘integrity’, even if the amount concentrations of characteristic constituents are not exactly same for different samples, hence ‘fuzziness’. [2] Therefore, HM extracts should be considered holistically, and the model of using only one or two marker components should not be considered for HM products. However, in any polyherbal powder and its extract, there are hundreds of unknown components, many of them present only in trace amounts. Consequently, to obtain reliable fingerprints that represent pharmacologically active and chemically characteristic components is not an easy or trivial task. [3] Recently increased quality requirements for herbal preparations raise new tasks to the science of pharmacy. Modern requirements are linked to better support of industrial production of traditional based multi-ingredient Phyto- preparations. In recent times the field of Chemometrics and multivariate analysis encompasses the interdisciplinary sciences of mathematics, analysis and chemistry to create results that are precise and specific, which is very essential in the field of medicine and has been widely implemented in the quality control herbal medicine. [4] On the other hand, many pharmaceutical companies have demand the automatic identification of powders from herbal plants. Due to lack of specialists, identification work is very tribal task. Taking Jayanta Kumar Maji et al. / International Journal of Pharma Sciences and Research (IJPSR) ISSN : 0975-9492 Vol. 7 No. 8 Aug 2016 325