With the advancement of computing resources, an increasing number of Neural Networks (NNs) are appearing for image detection and segmentation. However, these methods usually accept as input a RGB 2D image. On the other side, Light Detection And Ranging (LiDAR) sensors present a different number of horizontal layers, making those with a large number more expensive, but able to obtain data that is similar to that obtained by traditional low-resolution RGB cameras. With the purpose of whether or not is possible to perform object segmentation on pseudo-RGB images obtained from a large number of horizontal layer LiDAR sensor, a new dataset has been generated. This dataset combines the information given by the LiDAR sensor into a Spherical Range Image (SRI), concretely the reflectivity, near infrared and signal intensity 2D images. These images are then fed into instance segmentation NNs. These NNs segment the cars that appear in these images, having as a result Bounding Box (BB) and mask precision of 88% and 81.5% respectively with You Only Look Once (YOLO)-v8 large. By using this segmentation NN, some trackers have been applied so as to follow each car segmented instance along a video feed, having great performance in real world experiments. Results show that even with a small dataset, NNs are able to learn how to segment cars in pseudo-RGB images, leaving only the pending task of providing more samples so as to improve accuracy.