In Peru, phytoparasite nematodes are a type of specimen that adheres to vegetable roots, preventing plants from absorbing the necessary nutrients to grow. In order to apply contingency and prevention plans, it is necessary to study the families to which the nematode pests belong by analyzing their physical characteristics with a microscope. The proposed system represents an alternative to the current methods for the classification of phytoparasite nematodes through image processing. This system is an algorithm oriented to detect the physical characteristics of nematodes in tropical fruit crops (width and length). The algorithm involves image acquisition of nematodes through a microscope, obtain the luminance component of the image, illumination correction, binarization by histogram, objects segmentation, discrimination by area, surface detection and calculation of the Euclidean distance to approximate the physical characteristics of specimens in a sample of nematodes. Favorable results were obtained in the detection of the characteristics of the Meloidogyne type II species. The results were validated from those obtained by clinical analysis of the specialists, achieving up to 85% of success.