Real-Time Motion Detection via Census Transform and Sub-Sampling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing systems based on general-purpose computers face inefficiencies in real-time motion detection due to the need for repetitive calculations and large memory accesses, making it difficult to process motion information from images captured by cameras in real-time, especially as image size and frequency increase.
Innovation Solution
An image processing apparatus and method utilizing a hardware-based approach with a sub-sampling module, census transform module, correlation calculation module, and motion detection module to extract and process motion information in real-time, employing sub-sampling interpolation methods and census transforms to compare correlation values between images, allowing for efficient detection of moving objects without relying on software or computer systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a general-purpose computer uses sequential processing to check all pixels for motion detection, then motion information can be extracted, but processing speed and real-time capability significantly deteriorate
Solution Approach 1:
The patent divides the image processing task into multiple parallel processing units, each handling specific regions or aspects of motion detection simultaneously. This segmentation allows the system to maintain high processing speed while preserving detection accuracy through distributed parallel computation rather than sequential pixel-by-pixel analysis.
Solution Approach 2:
The patent replaces the mechanical sequential processing approach of general-purpose computers with a hardware-based parallel processing architecture. This substitution enables simultaneous comparison of current and previous images across multiple processing channels, achieving real-time motion detection without the speed limitations of software-based sequential methods.
2Loss of information
If image size and number of images input per second are increased, then more comprehensive motion information is available, but data processing volume increases making real-time processing impossible
Solution Approach 1:
The patent performs preliminary processing operations such as image differencing and feature extraction in hardware circuits before full analysis is needed. This preliminary action reduces the complexity of subsequent processing steps, enabling the system to handle larger image sizes and higher frame rates without increasing overall processing time.
Solution Approach 2:
The patent introduces parallel processing dimensions by implementing multiple processing channels that operate simultaneously on different aspects of the image data. This dimensional expansion allows the system to process comprehensive motion information from larger images at higher rates by distributing the computational load across multiple concurrent processing paths.
3Productivity
If a hardware-based method is used for real-time motion detection, then processing speed and real-time capability improve, but system complexity and implementation difficulty increase
Solution Approach 1:
The patent designs hardware processing units that can perform multiple functions within a single integrated circuit. These universal processors can handle various image processing operations including differencing, correlation calculation, and motion vector generation, reducing the need for separate dedicated hardware for each function and thereby managing system complexity while maintaining real-time capability.
Solution Approach 2:
The patent uses simplified hardware models and modular circuit designs that can be replicated and instantiated multiple times in parallel. By creating copies of basic processing units rather than designing complex monolithic circuits, the system achieves high processing speed through parallelism while keeping individual unit complexity manageable and design implementation feasible.
Data Source
AI summary
An image processing apparatus and method for real time motion detection is provided. In the apparatus, a sub-sampling module receives and sub-samples a current image and a plurality of previous images and a census transform module performs census transform on each of the sub-sampled images to obtain a census vector. A correlation calculation module calculates and compares correlation values between the current image and the plurality of previous images and detects a region having highest correlation. A motion detection module tracks positions of pixels corresponding to the region having the highest correlation to detect motion information in the images. The image processing apparatus and method can obtain, in real time, the direction and speed of an object that is in motion in each image.