The present work aims to estimate the length and width of the seeds, through the analysis of digital images and to validate the methodology through statistical data. To estimate the length and width of seeds via image analysis, 50 M.maripa seeds were used. The seeds were arranged in a decreasing way from 1 to 25 diagonally, on a matte black fabric, on a phenotyping platform, with a Sony Hd Avchd progressive digital Gps camera, coupled at a height of 50 cm. The images were captured by the camera in automatic mode, without flash, automatic ISO speed, in an RGB system and with a size above 2Mb. Soon after the seeds passed the traditional evaluation method with the aid of a digital caliper, measured in terms of length and width. The images were analyzed with the ImageJ software. Statistical analyzes were performed with the aid of the Sigmaplot program. The results of the length and width of the seeds of Inajá seen by the caliper and digital image via camera, were very distant, showing high dispersion and low correlation r = 0.4037 and R² = 0.1629 for length and r = 0.2985 and R² = 0.0891, showing that the Compared data had little similarity. The error between both the methodology was considered low: 3,81063 and 3,769 for the variables of seed length and width. The method of analysis by digital image and caliper for estimating the length and width of Maximiliana maripa seeds showed a low correlation between the two methodologies. The use of image analysis to estimate the length and width of M. maripa seeds is not indicated.

Castanheira-do-Brasil is a semi-deciduous species which belongs to forests made up of vegetation specific to the Amazon Forest biome. The objective of this work is to perform the leaf and adult morphometry of B. excelsa by the traditional method and to compare with image analysis processed using the ImageJ program. 30 images of chestnut trees were captured using a Sony GPS Avchd progressive câmera, in a commercial plantation, located between the coordinates 9.89974ºS and 56.32221ºW. The images were analyzed for total height, canopy width (DC), chest diameter height (DBH). All measurements were made using the ImageJ program. Statistical analyzes were performed using the Sigmaplot program. Leaves of B. excelsa were collected at random and measured for length and width manually with the aid of a ruler. The leaf area of these 50 leaves was estimated by the LI-COR® leaf area integrator. For the analysis by means of digital image, 50 photos were taken with the aid of a Sony GPS Avchd progressive camera, using a 50 cm high Phenotyping platform with a predefined scale. The images of the chestnut leaves were analyzed for length (CP) and width (L) in cm. All measurements were performed using the ImageJ program and the analyzes with the Sigmaplot program. The data obtained through digital images can be seen that the plants averaged 26.04 meters in height, 12.87 meters in canopy width and 0.80m in diameter at breast height. The relationship between the canopy width and the DAP allows the calculation of the maximum number of trees per hectare. The data obtained from the sum of the average length of chestnut leaves by the traditional method were 1575.1 cm, very close to the values found for the digital image method via a 1562.39 cm camera. For the variable leaf width, the sum via digital caliper was 547.1 cm and digital image 541.05 cm. The result obtained in this study is possible to state that there are significant relationships between all variables analyzed for the species B. excelsa. Where the ratio of canopy growth to increased plant height is proportionally equivalent.