An improved algorithm for object detection based on deep learning
Ke Li · 2023
The goal of object detection is to recognize the position and category of all objects in an image, allowing for machine vision understanding. Many approaches have been developed to solve this problem, primarily based on convolutional image feature extraction and deep learning algorithms. However, the current algorithms lack robustness and often have high error rates and a reliance on numerous parameters, making them difficult to apply in practical settings. To address these limitations, this paper presents an improved object detection method based on the R-CNN model. This method leverages the target feature matching module in the R-CNN model to accurately extract features through similarity calculation. Analysis showed that this improved method outperforms traditional models in terms of detection efficiency, especially in complex environments. The proposed method serves as a valuable reference for further research in related fields.