摘要
由于人脸皮沟的结构复杂且多样,常规的图像检测技术难以准确区分不同的皮沟区域,影响检测的可靠性。为此,研究设计一种基于深度视觉传感的人脸皮沟检测算法。利用色彩直方图提取图像目标,判断检测图像的颜色量化分布关系,计算色彩比例,调整图像的纹理参数,由此获取增强后的人脸皮沟检测图像,生成检测特征,提取高斯函数,计算中心偏移量,提取人脸皮沟检测特征,生成人脸皮沟检测优化算法,从而完成人脸皮沟检测。实验结果表明,该算法的JSI检测样本集相似度均值较高,成功优化了图像的直方图分布,减少了像素偏差,显著提升了图像的整体质量。
Because the structure of human facial sulcus is complex and diverse,it is difficult to accurately distinguish different sulcus areas by traditional image detection technology,which affects the reliability of detection.Therefore,the study designs a face furrow detection algorithm based on depth vision sensing.Using color histogram to extract image target,judge the color quantization distribution relationship of the detected image,calculate the color ratio,and adjust the texture parameters of the image,so as to obtain an enhanced facial groove detection image,generate a detection feature extraction Gaussian function,calculate the center offset,extract the facial groove detection features,and generate an optimization algorithm for facial groove detection,thus completing the facial groove detection.The experimental results show that the average similarity of JSI detection sample set of the algorithm is high,the histogram distribution of the image is successfully optimized,the pixel deviation is reduced,and the overall image quality is significantly improved.
作者
王利
WANG Li(Private Hualian College,Guangzhou 510663,China)
出处
《传感器世界》
2024年第10期32-38,共7页
Sensor World
