In the last years, the use of multispectral images in different applications has been increasing due to access to industrial multispectral cameras that can analyze the composition of different materials. In another scenario, the use of high-dynamic-range (HDR) imaging is not only related to different exposure times for analyzing the light but also to produce pretty pictures thanks to the access to portable cameras with HDR capabilities, like in the smartphones. In this work, we present a first approach to produce HDR images using nine spectral images acquired using a novel prototype of a multispectral camera using nanowires sensors. The methods analyzed were multiple linear regression, support vector machines, and the k-nearest neighbors algorithm. The approach is tested using a Macbeth chart of 24 colors and real, natural objects.