算法
稳健性(进化)
像素
计算机科学
颜色恒定性
转化(遗传学)
GSM演进的增强数据速率
人工智能
计算机视觉
图像(数学)
化学
生物化学
基因
标识
DOI:10.1109/iaeac50856.2021.9390872
摘要
In order to solve the problems of poor robustness and low real-time performance of traditional algorithms for locomotive number location, a vehicle number location method based on Retinex image enhancement and improved stroke width transformation algorithm is proposed. Retinex algorithm is used to enhance the image edge information, and the edge detection and maximum stable extremum region algorithm are used to roughly locate the train number. Then, the improved stroke width transformation algorithm is used to calculate the stroke width and stroke gradient, and the threshold value is set to limit the search range of pixels to avoid global search and improve the timeliness of the algorithm. The experimental results show that compared with the existing algorithms, the positioning accuracy of the proposed algorithm is improved by 4.72%, the time complexity is reduced by 45.03% compared with the traditional stroke width transformation algorithm, and it also shows good positioning effect when applied to other scenes.
科研通智能强力驱动
Strongly Powered by AbleSci AI