Frequency Domain HOG Feature Extraction for Image Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image detection and recognition methods using Histogram of Oriented Gradient (HOG) features are limited in detection ratio and accuracy due to the spatial domain processing of image pixels.
Innovation Solution
The method involves segmenting an image into blocks and cells, transforming pixels from the spatial domain to the frequency domain using Discrete Cosine Transform (DCT) or Discrete Fourier Transform (DFT), and identifying HOG features in the frequency domain to enhance detection and recognition accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If HOG feature extraction is performed in the spatial domain using conventional methods, then the process is straightforward and computationally simple, but the detection ratio and accuracy are limited
Solution Approach 1:
The patent transforms the image data from spatial domain to frequency domain using DCT or DFT, changing the fundamental parameter space in which HOG features are extracted. This parameter transformation enables more accurate gradient calculations by leveraging frequency domain properties, directly resolving the contradiction between detection accuracy and processing complexity.
Solution Approach 2:
The patent replaces the conventional spatial domain gradient calculation mechanism with a frequency domain-based approach. By using DCT/DFT transforms and operating in the frequency domain, the system substitutes the traditional mechanical gradient computation with a transformed domain approach that achieves superior accuracy.
2Productivity
If image processing is performed in the spatial domain, then the computational process is simpler, but the detection ratio is insufficient
Solution Approach 1:
The patent changes the domain parameter from spatial to frequency domain, enabling improved detection ratio through frequency-based gradient analysis. This parameter transformation allows the system to capture more discriminative features that are not apparent in the spatial domain, thereby increasing detection ratio despite increased processing complexity.
Data Source
AI summary
A method for identifying a feature of an image is provided. The method includes: segmenting an image into a plurality of blocks, wherein each block includes a plurality of cells; transforming pixels of each cell from a spatial domain to a frequency domain; and identifying a Histogram of Oriented Gradient (HOG) feature of the image in the frequency domain.