Semantic segmentation of objects in an image

In the previous section, we learned about performing segmentation on top of an image where the image contained only one object. In this segmentation, we will learn about performing segmentation so that we are able to distinguish between multiple objects that are present in an image of a road.

Getting ready

The strategy that we'll adopt to perform semantic segmentation on top of images of a road is as follows:

1. Gather a dataset that has the annotation of where the multiple objects within an image are located:
   • A sample of the semantic image looks as follows:

2. Convert the output mask into a multi dimensional array where there are as many columns as the number of all possible...