摘要 为提高边缘图像识别精度,设计一种多空间分辨率自适应特征融合边缘图像识别方法。首先对图像进行去噪处理,然后采用多空间分辨率自适应特征融合算法计算图像的特征值,对图像分割,最后分析图像的特征分布情况,实现了多空间分辨率自适应特征融合边缘图像识别。实验结果表明,此次研究的边缘图像识别方法不仅提高了边缘图像识别精度与信噪比,信噪比最高为64 dB,偏离度最低为2%,还减少了边缘图像识别时间,最低可在2.5 min时识别完毕,证明了多空间分辨率自适应特征融合边缘图像识别方法的有效性。 In order to improve the accuracy of edge image recognition,a multi-resolution adaptive feature fusion edge image recognition method is designed.First,the image is denoised,and then the feature value of the image is calculated by the multi-spatial resolution adaptive feature fusion algorithm,the image is segmented,and finally the feature distribution of the image is analyzed to realize the multi-spatial resolution adaptive feature fusion edge image recognition.The experimental results show that the edge image recognition method not only improves the edge image recognition accuracy and signal-to-noise ratio,the highest signal-to-noise ratio is 64 dB,and the minimum deviation is 2%,but also reduces the edge image recognition time.The minimum recognition time can be completed in 2.5 min,which proves the effectiveness of multi spatial resolution adaptive feature fusion edge image recognition method.
机构地区 广东邮电职业技术学院
出处 《自动化与仪器仪表》 2021年第4期48-51,55,共5页 Automation & Instrumentation
基金 2020年广东邮电职业技术学院质量工程教科研项目(No.202014,202037)。
关键词 多空间分辨率 自适应特征融合 边缘图像 识别 阈值 multi-spatial resolution adaptive feature fusion edge image recognition threshold
