Hello, trying to understand which strategy works best here.
I am using the mapping toolbox for a research project, which includes the calculation of the average distance/time for traveling from a number of polygonal ROIs (e.g. area of cities) to a given destination. This will in turn provide a measurement for the average cost that residents in the ROI have to endure in order to travel to the destination.
My script goes:
Consider one boundary polygon and geolocate it.
Build a grid of equally spaced geopoints within the polygon.
For each point in step 2, query the OSRM routing engine to the destination and store the response.
Average the distances/times.
I don’t like the current version of step 4, as the algorithm is applying the same weight to points in downtown and in the deep countryside. I would like to achieve a weighted average, where the weights represent a measure of the probability that someone lives nearby a given query point. This basically means detecting residential areas.
I’m not sure how to tackle this, as I’ve never worked with image processing (I’m assuming that’s the way, but I could be wrong). I thought about using geobasemap in ‘streets-light’ mode, and process the amount of pure white color in the picture, which would represent a measure of how much public road exists in a given portion of my ROI. The more roads are present, the less likely it is that we’re in the deep countryside.
I’m pretty sure there are more appropriate workarounds. Has anyone encountered a similar problem? Thank you very much.Hello, trying to understand which strategy works best here.
I am using the mapping toolbox for a research project, which includes the calculation of the average distance/time for traveling from a number of polygonal ROIs (e.g. area of cities) to a given destination. This will in turn provide a measurement for the average cost that residents in the ROI have to endure in order to travel to the destination.
My script goes:
Consider one boundary polygon and geolocate it.
Build a grid of equally spaced geopoints within the polygon.
For each point in step 2, query the OSRM routing engine to the destination and store the response.
Average the distances/times.
I don’t like the current version of step 4, as the algorithm is applying the same weight to points in downtown and in the deep countryside. I would like to achieve a weighted average, where the weights represent a measure of the probability that someone lives nearby a given query point. This basically means detecting residential areas.
I’m not sure how to tackle this, as I’ve never worked with image processing (I’m assuming that’s the way, but I could be wrong). I thought about using geobasemap in ‘streets-light’ mode, and process the amount of pure white color in the picture, which would represent a measure of how much public road exists in a given portion of my ROI. The more roads are present, the less likely it is that we’re in the deep countryside.
I’m pretty sure there are more appropriate workarounds. Has anyone encountered a similar problem? Thank you very much. Hello, trying to understand which strategy works best here.
I am using the mapping toolbox for a research project, which includes the calculation of the average distance/time for traveling from a number of polygonal ROIs (e.g. area of cities) to a given destination. This will in turn provide a measurement for the average cost that residents in the ROI have to endure in order to travel to the destination.
My script goes:
Consider one boundary polygon and geolocate it.
Build a grid of equally spaced geopoints within the polygon.
For each point in step 2, query the OSRM routing engine to the destination and store the response.
Average the distances/times.
I don’t like the current version of step 4, as the algorithm is applying the same weight to points in downtown and in the deep countryside. I would like to achieve a weighted average, where the weights represent a measure of the probability that someone lives nearby a given query point. This basically means detecting residential areas.
I’m not sure how to tackle this, as I’ve never worked with image processing (I’m assuming that’s the way, but I could be wrong). I thought about using geobasemap in ‘streets-light’ mode, and process the amount of pure white color in the picture, which would represent a measure of how much public road exists in a given portion of my ROI. The more roads are present, the less likely it is that we’re in the deep countryside.
I’m pretty sure there are more appropriate workarounds. Has anyone encountered a similar problem? Thank you very much. image processing, matlab, figure, image analysis, geobasemap MATLAB Answers — New Questions

​