Method and system for facilitating real time detection of linear infrastructural objects by aerial imagery
This disclosure relates generally to visual inspection systems, and more particularly to a method and system for facilitating real time detection of linear infrastructural objects in aerial imagery. In one embodiment, a background suppression technique is applied to one or more hardware processors to a HSV image. Further, a mean shift filtering technique is applied to the hardware processors to find a peak of a confidence map and then a gradient image generation is performed for a plurality of edges of the image. A seed point pair along a middle cut portion of a linear feature of the HSV image to identify one or more boundaries of the seed point pair is extracted and then a contour growing approach to detect the boundaries of the linear feature is initiated. Lastly, one or more false positives are removed by using a rigidity feature, the rigidity feature being equivalent to the total sum of gradient orientations.
1. A processor-implemented method for facilitating real time detection of at least one linear infrastructural object by aerial imaging, said method comprising:
applying, via one or more hardware processors, a background suppression technique to a HSV image, wherein said HSV image is first converted to a grey scale image and a binary image;
implementing, via the one or more hardware processors, a mean shift filtering technique to find a peak of a confidence map by using a color histogram of said HSV image;
performing, via the one or more hardware processors, a gradient image generation for a plurality of edges of said HSV image using a sobel function;
extracting via the one or more hardware processors a seed point pair along a middle cut portion of a linear feature of the HSV image to identify one or more boundaries of the seed point pair;
initiating via the one or more hardware processors a contour growing approach to detect said one or more boundaries of the linear feature; and
removing via the one or more hardware processors one or more false positives by using a rigidity feature, the rigidity feature being equivalent to the total sum of gradient orientations.
2. The processor implemented method of claim 1 , further comprising constructing a three dimensional feature space to locate a plurality of high fidelity conductors of said at least one infrastructural object.
3. The processor implemented method of claim 1 , wherein a plurality of linear features are detected by tracking a boundary of said infrastructural object in linear space, by applying said background suppression technique.
4. The processor implemented method of claim 1 , wherein the seed point pair is constructed by using a second set of candidate seed points, by a side-facing camera.
5. The processor implemented method of claim 1 , wherein the seed point pair is detected from a bottom horizontal line, by a front facing camera.
6. A system for facilitating real time detection of at least one infrastructural object by aerial imaging, said system comprising:
a memory storing instructions;
a hardware processor coupled to said memory, wherein said hardware processor is configured by said instructions to:
apply a background suppression technique, wherein a HSV image is first converted to a grey scale image and a binary image;
implement a mean shift filtering technique to find a peak of a confidence map by using a color histogram of said HSV image;
perform a gradient image generation for a plurality of edges of the HSV image using a sobel function;
extract a seed point pair along a middle cut portion of a linear feature of the HSV image to identify one or more boundaries of the seed point pair; and
initiate a contour growing approach to detect said one or more boundaries of the linear feature.
7. The system of claim 6 , wherein said hardware processor is configured to construct a three dimensional feature space to locate a plurality of high fidelity conductors of said at least one infrastructural object.
8. The system of claim 6 , wherein said hardware processor is configured to detect a plurality of linear features by tracking a boundary of said infrastructural object in linear space, by said background suppression technique.
9. The system of claim 6 , wherein said hardware processor is configured to construct the seed point pair by using a set of candidate seed points.
10. The system of claim 6 , wherein said real time detection of at least one infrastructural object is enabled by using at least one of a front facing camera, a rear facing camera, and a side facing camera.