Crop phenotyping is a desirable task in crop characterization since it allows the farmer to make early decisions, and therefore be more productive. This research is motivated by the generation of tools for rice crop phenotyping within the OMICAS research ecosystem framework. It proposes implementing the image process- ing technologies and artificial intelligence technics through a multisensory approach with multispectral information. Three main stages are covered: (i) A segmentation approach that allows identifying the biological material associated with plants, and the main contri- bution is the GFKuts segmentation approach; (ii) a strategy that allows the development of sensory fusion between three different cameras, a 3D camera, an infrared multispectral camera, and a thermal multispectral camera, this stage is developed through a complex object detection approach; and (iii) the characterization of a 4D model that generates topological relationships with the information of the point cloud, the main contribution of this strategy is the improvement of the point cloud captured by the 3D sensor, in this sense, this stage improves the acquisition of any 3D sensor. This research presents a development that receives information from multiple sensors, especially infrared 2D, and generates a single 4D model in geometric space [X, Y, Z]. This model integrates the color information of 5 channels and topological information, relating the points in space. Overall, the research allows the integration of the 3D information from any sensor\technology and the multispectral channels from any multispectral camera, to generate direct non-invasive measurements on the plant.