Abstract: The present invention concerns a zeolitic adsorbent agglomerate comprising at least one zeolite of faujasite type comprising barium and/or potassium, of porosity between 25% and 45%, and having a standard deviation ? of crystal size distribution in said agglomerate of less than 0.30 ?m. The invention also concerns the use of the zeolitic adsorbent agglomerate for the separation of hydrocarbon mixtures, and the process for separating hydrocarbon mixtures using said zeolitic adsorbent agglomerate.

Abstract: The present invention relates to a method of detecting a geological constituent (cge) of a rock sample, by use of a machine learning method (ALG) trained with a limited number of training images (IAP), through segmentation (SEG) and partitioning (DEC) of training images (IAP) into multiple patches.

Abstract: The present invention relates to a method of detecting a geological constituent (cge) of a rock sample, by use of a machine learning method (ALG) trained with a limited number of training images (IAP), through segmentation (SEG) and partitioning (DEC) of training images (IAP) into multiple patches.

Abstract: The invention is a method for the segmentation of an object's image obtained by three-dimensional reconstruction. Based on an object's image represented by a finite number of intensity levels, the image is reconstructed based on projections acquired from different angles around the object, a distribution representative of the spread of the intensity levels in the image along a predetermined direction is determined, for at least one angle. Next, at least one intensity threshold is determined by minimizing an objective function estimating the difference between the actual measurement and a simulated measurement for the angle being considered with the simulated measurement being dependent on the threshold and the distribution. Next, the image is segmented by grouping the intensity levels into at least two classes defined by the threshold. The invention has a particular application to characterization of supports of catalysts for petrochemical refining.

Abstract: The invention is a method for the segmentation of an object's image obtained by three-dimensional reconstruction. as Based on an object's image represented by a finite number of intensity levels, the image is reconstructed based on projections acquired from different angles around the object, a distribution representative of the spread of the intensity levels in the image along a predetermined direction is determined, for at least one angle. Next, at least one intensity threshold is determined by minimizing an objective function estimating the difference between the actual measurement and a simulated measurement for the angle being considered with the simulated measurement being dependent on the threshold and the distribution. Next, the image is segmented by grouping the intensity levels into at least two classes defined by the threshold. The invention has a particular application to characterization of supports of catalysts for petrochemical refining.