18053184. METHOD OF OPERATING DYNAMIC VISION SENSOR SYSTEM simplified abstract (SAMSUNG ELECTRONICS CO., LTD.)
Contents
METHOD OF OPERATING DYNAMIC VISION SENSOR SYSTEM
Organization Name
Inventor(s)
METHOD OF OPERATING DYNAMIC VISION SENSOR SYSTEM - A simplified explanation of the abstract
This abstract first appeared for US patent application 18053184 titled 'METHOD OF OPERATING DYNAMIC VISION SENSOR SYSTEM
Simplified Explanation
The patent application describes a method for operating a dynamic vision sensor system. Here are the key points:
- The system obtains event signals from dynamic vision sensor pixels, which measure changes in light over a specific time period.
- Based on these event signals, the system obtains original image data, which consists of image pixels corresponding to the dynamic vision sensor pixels.
- The system then binarizes the pixel values of the image pixels to obtain binary image data.
- From the binary image data, the system defines pixel groups.
- Finally, the system selects effective groups from these pixel groups, which represent the movement of an object.
Potential applications of this technology:
- Object tracking and recognition in surveillance systems
- Motion detection in autonomous vehicles
- Gesture recognition in human-computer interaction
Problems solved by this technology:
- Traditional image sensors capture and process every pixel, leading to high computational requirements and power consumption.
- Dynamic vision sensors only capture changes in light, reducing computational load and power consumption.
- The method described in the patent application further simplifies the processing by selecting only the relevant pixel groups.
Benefits of this technology:
- Lower computational requirements and power consumption compared to traditional image sensors.
- Improved efficiency in object tracking and recognition tasks.
- Faster response time due to the selective processing of relevant pixel groups.
Original Abstract Submitted
A method of operating a dynamic vision sensor system includes: obtaining event signals from a plurality of dynamic vision sensor pixels over a predetermined time period, wherein the event signals correspond to a measured change in light; obtaining original image data based on the event signals output by the plurality of DVS pixels and including a plurality of image pixels, wherein the plurality of image pixels respectively correspond to the plurality of dynamic vision sensor pixels; obtaining binary image data by binarizing pixel values of the plurality of image pixels; defining a plurality of pixel groups from the plurality of image pixels in the binary image data; and selecting a plurality of effective groups from among the plurality of pixel groups that represent the movement of an object.
