Image Sensor ROI Processing for Low-Bandwidth Motion Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing CMOS image sensors face challenges in capturing high-speed, high-resolution images with large bandwidth and power consumption, particularly in applications requiring real-time motion detection.
Innovation Solution
An image sensor integrated with a processing-in-sensor (PIS) circuit that generates frame-difference signals and identifies regions of interest (ROI) without external processor computing, reducing power consumption and latency by processing images internally.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high frame-rate images with high spatial resolution are captured, then image quality is improved, but bandwidth and power consumption increase
Solution Approach 1:
The patent divides the image processing task by identifying and processing only the region of interest (ROI) rather than the entire image. The circuit determines motion detection regions and processes only those areas, significantly reducing the data volume that requires transmission and external processing, thereby lowering power consumption while maintaining image quality in the critical motion areas.
Solution Approach 2:
The patent extracts and processes only the essential information (motion detection regions) from the full image data. By using frame difference signals and comparing them against threshold values, the circuit identifies and outputs only the ROI containing motion information, eliminating unnecessary data transmission and processing externally, thus reducing overall power consumption.
2Measurement precision
If high frame-rate images with high spatial resolution are captured, then image quality is improved, but bandwidth requirements increase
Solution Approach 1:
The patent segments the image processing by focusing only on motion detection regions rather than processing the entire image. This division allows the system to transmit and process only the essential motion information, reducing the data volume and bandwidth requirements while preserving image quality in the critical motion areas.
Solution Approach 2:
The patent extracts only the necessary motion detection regions from the full image data and transmits/processes only that extracted information. By using frame difference signals and threshold comparison to identify ROIs, the system eliminates redundant data transmission, significantly reducing bandwidth requirements while maintaining image quality where it matters most.
3Measurement precision
If frame-difference signals are processed to generate ROI images, then motion detection accuracy is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary processing of frame-difference signals within the sensor itself, generating motion detection regions before the data reaches external processors. By pre-identifying ROIs using threshold comparison of frame differences, the system prepares and filters data in advance, reducing the processing burden and time required for subsequent external analysis while maintaining motion detection accuracy.
Solution Approach 2:
The sensor performs self-processing by generating and processing frame-difference signals internally to identify motion detection regions. This self-service capability allows the sensor to prepare and filter data before external processing, reducing overall processing time while maintaining accuracy through the integrated frame difference comparison and threshold evaluation performed within the sensor itself.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables high-speed, low-latency, and energy-efficient motion detection by generating ROI images directly within the sensor, reducing the need for external processing and minimizing bandwidth requirements.
Implementation Method 1
a pixel array and an analog-to-digital converter generate raw images of frames and a processor detects moving objects by calculating motion detection values throughout the raw images
Data Source
AI summary
A method is provided to operate an image sensor, including the following steps: comparing a first threshold with multiple first frame-difference signals to generate a first region of interest (ROI) address, wherein each of multiple first frame-difference signals is generated by a corresponding pixel, in multiple pixels, operating during a first time frame and a second time frame, wherein multiple pixels are arranged in multiple columns and multiple rows; identifying edge columns and edge rows in the first ROI address; adjusting, according to the edge columns and the edge rows, the first ROI address to generate a second ROI address; and outputting a ROI image according to the second ROI address and multiple second frame-difference signals generated by multiple pixels operating during the second time frame and a third time frame after the second time frame.