光针轮廓扫描技术测量内螺纹曲面

分析了激光光针扫描在测量内螺纹曲面时的特性及影响精度的因素。根据光斑在被测面的噪声分布特点,采用中值滤波与小波阈值去噪相结合的综合滤波算法对光斑图像进行了去噪处理,其中中值滤波可以有效地消除光斑图像中的脉冲噪声,而基于小波变换的阈值去噪算法可以消除图像中的高斯噪声。通过综合滤波算法对光斑图像进行处理,可以有效消除激光光斑噪声,消除图像噪声对测量精度的影响,同时保持了光斑图像的细部特征。根据激光光斑图像在倾斜被测物体表面光强分布特点,设计了近似圆拟合算法来确定光斑中心,该算法相对传统算法具有计算量小、计算效率高,适合在线检测的特点,可以克服测量面倾斜对激光光斑中心的影响。利用采用上述方法的坐标机激光光针扫描系统,对标准内螺纹量规进行测量,对比应用前后的测量结果,螺距精度提高了0.48 mm,中径测量精度提高了0.44 mm,牙型角精度提高0.57°。

Measurement of inside screw curve surface by laser probe

The characters and factors were analyzed when the laser probe measurement was used in inside screw surface. According to the noise distribution of the laser speckle on the measured plane, a synthesis filter algorithm combined the median filter and the wavelet threshold denoising method was designed to eliminate noise of the image,in which median filter can effectively wipe off the impulse noise in the speckle image, and the threshold denoising algorithm based on wavelet transform can eliminate Gaussian noise of the image. The algorithm is able to remove the noise influence on measurement accuracy and maintain the image individual character. According to the intensity distribution of the laser speckle image on the declining object surface, the circle-fitting algorithm was proposes to overcome the affect of the laser speckle center shift caused by the slope surface to locate the center of the laser speckle image at the sub-pixel level. The algorithm has the features of high efficiency, less calculation compared with the traditional ones and is more suitable for testing on line. Laser probe scanning measurement base on coordinate machine using the above techniques was designed to measure an internal screw standard gauge. The results show that the internal screw-pitch measurement accuracy, pitch diameter of thread and the thread angle accuracy increase 0.48 mm, 0.44 mm and 0.57°,respectively.