Where PANORAMAID = panoramic id found in the metadata, XPOS = column position, and YPOS = row position. The columns and rows need to be calculated as an integer datatype. Next an iterative procedure is used to establish a list of the Google Maps REST API urls needed to obtain all the tiles. The second step involves determining the number or tile rows and columns that will be requested for a given zoom level. First, we need to determine the output resolution for an input zoom level. The first step involves finding the metadata associated with the nearest panoramic image to a provided latitude and longitude. A call is made to Google Maps REST API to obtain metadata in JSON format. Simply replace “LAT” with your latitude and “LNG” with your longitude.Ī request is made using python requests to obtain the JSON metadata and the metadata is converted into a Python dictionary to obtain the values to the fields of interest, such as the panorama identifier label assigned by Google for the unique panoramic image, and overall panoramic image height and width.

Reconstructing the tiles into a complete image.īefore getting to the script, I will describe the overall process used by the script to obtain the spherical imagery. Inputting a zoom level to adjust output resolution of the image (1 is lowest resolution, 5 is highest resolution).Ĥ.

Obtaining metadata associated with the specific image, specifically the unique panoramic identification label and full image pixel resolution height and width.Ģ. I was recently exploring how to utilize Google Street View depth maps in combination with spherical imagery to reconstruct basic point clouds of city environments. I came across an article describing the process. I created a Python script to compose a panorama using Google Maps REST API using the methodology presented in the article. My script demonstrates:ġ. You can open the raw images in any spherical (360° panoramic) image viewer or use in a VR environment. The purpose of this article is to demonstrate how to obtain complete spherical images captured in Google Street View using Python.

Earth view 360