US 2019/0197724 A1 · CLUTTERED BACKGROUND REMOVAL FROM IMAGERY FOR OBJECT DETECTION

Abstract

Embodiments herein describe tracking the location and orientation of a target in a digital image. In an embodiment, this tracking can be used to control navigation for a vehicle. In an embodiment, a digital image can be captured by a visual sensor is received. A first array including a plurality of binary values related to the pixel velocity of a first plurality of pixels in the digital image as compared to corresponding pixels in a first one or more prior digital images can be generated. A second array including a plurality of values related to the standard deviation of pixel intensity of the first plurality of pixels in the digital image as compared to corresponding pixels in a second one or more prior digital images can be further generated. A plurality of thresholds relating to the values in the second array can be determined. A plurality of target pixels and a plurality of background pixels can be identified in the digital image, based on the first array, the second array, and the plurality of thresholds. A binary image related to the digital image, based on the identified plurality of target pixels and the identified plurality of background pixels, and identifying at least one of a location and an orientation of the target in the digital image based on the binary image, can be generated. In an embodiment, a command can be transmitted to a navigation system for a vehicle, to assist in navigating the vehicle toward the target, based on the identified at least one of a location and an orientation of the target.

Claims20

Dependent claims

A method for controlling navigation of a vehicle by tracking a location and orientation of a target in a digital image, the method comprising:

receiving a digital image captured by a visual sensor;
generating a first array comprising a plurality of binary values related to a pixel velocity of a first plurality of pixels in the digital image as compared to corresponding pixels in a first one or more prior digital images;
generating a second array comprising a plurality of values related to a standard deviation of pixel intensity of the first plurality of pixels in the digital image as compared to corresponding pixels in a second one or more prior digital images;
determining a plurality of thresholds relating to the values in the second array;
identifying a plurality of target pixels and a plurality of background pixels in the digital image, based on the first array, the second array, and the plurality of thresholds;
generating a binary image related to the digital image, based on the identified plurality of target pixels and the identified plurality of background pixels;
identifying at least one of a location and an orientation of the target in the digital image based on the binary image;
and
transmitting a command to a navigation system for the vehicle, to assist in navigating the vehicle toward the target, based on the identified at least one of the location and the orientation of the target.

6 more elements

A method for tracking a location and orientation of a target in a digital image, the method comprising:

receiving a digital image captured by a visual sensor;
generating a first array comprising a plurality of binary values related to a pixel velocity of a first plurality of pixels in the digital image;
generating a second array comprising a plurality of values related to a standard deviation of pixel intensity of the first plurality of pixels in the digital image;
identifying a plurality of target pixels and a plurality of background pixels in the digital image, based on the first array and the second array;
and
identifying at least one of a location and an orientation of the target in the digital image based on the identified plurality of target pixels and a plurality of background pixels.

3 more elements

A system comprising:

a processor;
and
a memory storing a program, which, when executed on the processor, performs an operation, the operation comprising:
receiving a digital image captured by a visual sensor;
generating a first array comprising a plurality of binary values related to a pixel velocity of a first plurality of pixels in the digital image;
generating a second array comprising a plurality of values related to a standard deviation of pixel intensity of the first plurality of pixels in the digital image;
identifying a plurality of target pixels and a plurality of background pixels in the digital image, based on the first array and the second array;
and
identifying at least one of a location and an orientation of the target in the digital image based on the identified plurality of target pixels and a plurality of background pixels.

6 more elements

Dependencies are inferred from the claim text.

Cluttered background removal from imagery for object detection

Life of this patent

Cluttered background removal from imagery for object detection

Life of this patent

Abstract

Claims20