Vigour is an essential parameter in vineyard management. Highest grape quality potential is obtained when vine vigour is moderate. Excessive vigour induces high yield and shaded canopy, resulting in quality losses. Excessively low vigour reduces crop yield. Vine vigour is highly variable among vineyard blocks as well as inside vineyard blocks. The reasons for this variability are mainly soil type, soil depth and topography. Spatial representation of vine vigour (vigour maps) can provide a useful tool for vineyard management. Traditionally, vine vigour is estimated either by the measurement of pruning weight or by the measurement of leaf area, and particularly of secondary leaf area. These techniques are labour intensive and they cannot be implemented to map vine vigour at a reasonable cost. Vine vigour can be mapped by remote sensing, using vegetation indexes like NDVI (Normalized Differential Vegetation Index). In low density vineyards, row width varies with vine vigour. In these conditions vine vigour can be mapped by means of remote sensing without differentiating pixels from the canopy and the soil, given that no vegetation is present on the soil. In high density vineyards, vine canopy is severely trimmed and row width does not vary with vine vigour, making vigour mapping by remote sensing much more complex. It is only possible, when very high resolution images contain pixels largely inferior in size to the row with, i.e. inferior to 10 cm, in order to separate pixels from canopy and pixel from soil. We developed a semi-automatic method to transform high resolution images in vine vigour maps, by using only pixels from the canopy. Images are spatially referenced. Red and Near Infra Red images are combined to construct a NDVI image. An algorithm based on deformable templates was developed to find automatically the centres of the rows. NDVI is calculated on the pixels of the centres of the rows and on the two adjacent pixels on each side. All other pixels (soil pixels, mixed soil-vegetation pixels) are excluded. Anisotropic NDVI data is transformed in a NDVI map by means of a Geographic Information System (G.I.S.). The NDVI vigour maps can be used in a precision viticulture approach, to implement an intra-block modulation of the use of fertilizers or pesticide sprayings depending on the density of vine vegetation.