Back to results
US 11,900,643 B2Granted
Object detection method and object detection system
AssigneeHimax Technologies Limited
InventorsI-Wei Chen, Ta-Jen Kuo, Pin-Rong Chen
Life of this patent
Estimating the term…
Abstract
An object detection method, for detecting a target object, comprising: capturing at least two detection portions with a first aspect ratio from an input image with a second aspect ratio; confirming whether any object is detected in each of the detection portions and obtaining corresponding boundary boxes for detected objects; and wherein the first aspect ratio is different to the second aspect ratio.
Claims14
An object detection method, for detecting a target object, comprising:
- capturing at least two detection portions with a first aspect ratio from an input image with a second aspect ratio;
- confirming whether any object is detected in each of the detection portions and obtaining corresponding boundary boxes for detected objects;
- and
- removing or merging the boundary boxes, comprising:
- defining at least one filtering region in the input image;
- classifying the boundary box having an edge in the filter region as a candidate boundary box, and classifying the boundary box having no edge in the filter region as a maintained boundary box;
- and
- removing the candidate boundary box from the boundary boxes if
- AI
- MA
- is larger than a threshold value, wherein AI is the area of an intersection region of the candidate boundary box or the maintained boundary box, and MA is a minimum one of the areas of the candidate boundary box and the maintained boundary box;
- wherein the first aspect ratio is different to the second aspect ratio.
9 more elements
An object detection system, for detecting a target object, comprising:
- a partial image capturing device, configured to capture at least two detection portions with a first aspect ratio from an input image with a second aspect ratio;
- an object detector, configured to receiving the at least two detection portions with the first aspect ratio and then confirm whether any object is detected in each of the detection portions and to obtain corresponding boundary boxes for detected objects;
- and
- a filter, configured to remove or to merge the boundary boxes, by:
- defining at least one filtering region in the input image;
- classifying the boundary box having an edge in the filter region as a candidate boundary box, and classifying the boundary box having no edge in the filter region as a maintained boundary box;
- and
- removing the candidate boundary box from the boundary boxes if
- AI
- MA
- is larger than a threshold value, wherein AI is the area of an intersection region of the candidate boundary box or the maintained boundary box, and MA is a minimum one of the areas of the candidate boundary box and the maintained boundary box;
- wherein the first aspect ratio is different to the second aspect ratio.
9 more elements
Dependencies are inferred from the claim text.