Integral Image Construction Using Reduced Bit Depth Accumulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing methods for mobile devices require significant memory space to store integral images, which can lead to high costs and slow calculation speeds, and often result in overflow values when using fewer bits for integration operations.
Innovation Solution
An image processing method that uses a graphics processing unit to perform image integration by accumulating pixel values with a predetermined number of bits less than log2(W×H×2k), allowing for the construction of an integral image that requires less memory space while handling overflow values, and calculates the gray-level sum of a region of interest using integral image values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Volume of stationary object
If a predetermined number of bits less than log2(W×H×2k) is used for integration operation, then memory space requirement is reduced, but overflow values occur in integral values
Solution Approach 1:
The patent changes the bit depth parameter from the traditional log2(W×H×2k) bits to a reduced predetermined number of bits for storing integral values. This parameter change directly reduces memory space requirements while the system accepts and manages the resulting overflow conditions through appropriate handling mechanisms in the image processing pipeline.
2Speed
If integral image is stored in volatile memory, then calculation speed is improved, but memory cost increases
Solution Approach 1:
The patent uses only a portion of the available volatile memory by reducing the bit depth for storing integral values. Instead of allocating full log2(W×H×2k) bits per pixel, it uses a predetermined smaller number of bits, thereby utilizing volatile memory efficiently for fast access while minimizing the total memory quantity required and associated costs.
3Measurement precision
If more bits are used for integration operation, then precision of integral values is improved, but memory space requirement increases
Solution Approach 1:
The patent deliberately changes the bit depth parameter to a smaller predetermined value rather than using the maximum log2(W×H×2k) bits. This parameter optimization finds a balance point where the reduced precision is acceptable for the application while achieving significant memory space savings, demonstrating that full precision is not always necessary for practical image processing tasks.
Data Source
AI summary
The present invention provides an image processing method applied to a graphics processing unit including: receiving data of an image, wherein the image is a first rectangle constituted by a plurality of pixels, and the pixels of the image are represented by a plurality of image values with a predetermined bit depth in the data; using a predetermined number of bits to perform an accumulation of the image values of the pixels for constructing an integral image of the image, wherein the predetermined number of bits is less than log2(W×H×2k) number of bits, W is the width of the first rectangle, H is the length of the first rectangle, and k is the predetermined bit depth.