.png){kind=icon}
.png){kind=icon}
Return
Advanced HOG Research: Multi-Scale Partitioning HOG Algorithm
.png){kind=icon}
PRE
AI
DOI：10.1109/ACCESS.2025.3609976.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The Histogram of Oriented Gradients algorithm is a widely adopted technique in visual feature extraction, especially in object detection tasks. Traditionally, HOG descriptors rely on fixed-size cells and blocks to capture local gradient distributions. Although this technique has been proven effective in pedestrian detection, its single-scale structure limits its ability to represent features in changing spatial patterns. To address this limitation, this study introduces an improved feature extraction method that integrates multi-scale partitioning. Inspired by feature pyramid representation, the method captures image characteristics in different receptive fields by using multiple combinations of cell and block sizes. These features are aggregated to form a more discriminative and robust representation. Comprehensive experiments on multiple public datasets demonstrate that the proposed multi-scale HOG framework consistently outperforms standard HOG methods, providing remarkable improvements in both classification and detection accuracy. The results validate the effectiveness and adaptability of the method in diverse visual recognition scenarios.
Keywords:
Feature extraction
Principal component analysis
Vectors
Pedestrians
Histograms
Accuracy
Standards
Data mining
Partitioning algorithms
Visualization
Histogram of oriented gradients (HOG)
multi-scale feature extraction
feature fusion
image classification