Detection of objects using range information
View Patent ↗A system and method for detecting objects and background in digital images using range information includes receiving the digital image representing a scene; identifying range information associated with the digital image and including distances of pixels in the scene from a known reference location; generating a cluster map based at least upon an analysis of the range information and the digital image, the cluster map grouping pixels of the digital image by their distances from a viewpoint; identifying objects in the digital image based at least upon an analysis of the cluster map and the digital image; and storing an indication of the identified objects in a processor-accessible memory system.
1. A method for detecting objects and background in a digital image, and the method implemented at least in part by a data processing system and comprising the steps of:
receiving the digital image representing a scene;
identifying range information associated with the digital image, the range information including distances of pixels in the scene from a known reference location;
generating a cluster map based at least upon an analysis of the range information using a clustering algorithm, the cluster map grouping pixels of the digital image by their distances from a viewpoint;
identifying objects in the digital image based at least upon an analysis of the cluster map and the digital image, wherein the analysis of the cluster map and the digital image includes:
identifying edges in the digital image; and
refining the pixel groupings in the cluster map by comparing the identified edges with borders of the pixel groupings and removing pixels that are outside of the identified edges from the pixel groupings; and
storing an indication of the identified objects in a processor-accessible memory system.
2. The method of claim 1 , further comprising labeling the identified objects according to their distances from the viewpoint, and storing results of the labeling in a processor-accessible memory system.
3. The method of claim 1 , wherein the identified edges undergo a filtering process to remove insignificant edges prior to being compared with the borders of the pixel groupings.
4. The method of claim 1 , wherein the range information is received from a ranging camera that senses range information using visible light, inferred light, ultrasound, or laser light.
5. The method of claim 1 wherein the range information is received from a stereoscopic imaging system that determines range information by analyzing a plurality of images of the scene captured from different viewpoints.
6. The method of claim 1 , wherein the digital image is captured by a digital camera or a scanner.
7. The method of claim 1 , further comprising the step of identifying a viewpoint location.
8. The method of claim 7 , wherein the viewpoint location is the reference location.
9. The method of claim 1 , wherein the analysis of the cluster map includes the step of reducing noise from the cluster map.
10. The method of claim 9 , wherein the reducing of noise from the cluster map fills small holes and removes small clustered regions from the cluster map.
11. A method for detecting objects and background in a digital image, and the method implemented at least in part by a data processing system and comprising the steps of:
receiving the digital image representing a scene;
identifying range information associated with the digital image, the range information including distances of pixels in the scene from a known reference location;
generating a plurality of cluster groups based at least upon an analysis of the range information using a clustering algorithm, the cluster groups representing groupings of pixels in the digital image according to their distances from a viewpoint;
identifying objects in the digital image based at least upon an analysis of the cluster groups and the digital image, wherein the analysis of the cluster groups and the digital image includes:
identifying edges in the digital image; and
refining the cluster groups by comparing the identified edges with borders of the cluster groups and removing pixels that are outside of the identified edges from the cluster groups; and
storing an indication of the identified objects in a processor-accessible memory system.
12. A system for detecting objects in a digital image comprising:
a data processing system; and
a memory system communicatively connected to the data processing system, the memory system storing instructions configured to cause the data processing system to implement a method for detecting objects in a digital image, wherein the instructions comprise:
receiving the digital image representing a scene;
identifying range information associated with the digital image, the range information including distances of pixels in the scene from a known reference location;
generating a cluster map based at least upon an analysis of the range information using a clustering algorithm, the cluster map grouping pixels of the digital image by their distances from a viewpoint; and
identifying objects in the digital image based at least upon an analysis of the cluster map and the digital image, wherein the analysis of the cluster map and the digital image includes:
identifying edges in the digital image; and
refining the pixel groupings in the cluster map by comparing the identified edges with borders of the pixel groupings and removing pixels that are outside of the identified edges from the pixel groupings.
13. The system of claim 12 , wherein the range information is received from a ranging camera that senses range information using visible light, inferred light, ultrasound, or laser light.
14. The system of claim 12 wherein the range information is received from a stereoscopic imaging system that determines range information by analyzing a plurality of images of the scene captured from different viewpoints.
15. The system of claim 12 , wherein the digital image is captured by a digital camera or a scanner.