Quality Criteria for Determination of Seeds Nutmeg (Myristica fragrans Houtt) Based Texture Analysis Using Digital Image Processing Technology

Separation of nutmeg based on quality at the farm level is still not done. At the market level process to separate the whole seed and seed damage done by direct observation. The process has the disadvantage, among others, can not be done continuously and mixed results. Development of non-destructive method for separate nutmeg by class quality effectively and objectively indispensable. On image texture analysis can be used to differentiate the surface properties of an object in the image associated with the rough and smooth, also the specific properties of the surface roughness and smoothness criteria that characterize an object of an object. This study aims to analyze the texture characteristics of the object image nutmeg with image processing to determine the quality grade of nutmeg. The materials used are nutmeg derived from Ternate town of North Maluku with reference to defined quality standards in 2000 that divides Menegristek nutmeg into three quality classes ABCD, Rimpel and BWP. Determination of the quality criteria nutmeg done by the method of discriminant analysis. Texture characteristics extracted from the object image consisting of nutmeg contrast, correlation, energy, homogenity, entropy. The results showed significant parameter correlation and the entropy distinguish quality classes nutmeg with a degree of truth of 96,7%.