Eliminating the zero spectrum in Fourier transform profilometry using a two-dimensional continuous wavelet transform

In this paper, a filtering technique based upon two-dimensional continuous wavelet transform (2D-CWT) is used to eliminate the low frequency components of fringe patterns. The filtered fringe patterns are subsequently demodulated using a standard Fourier transform profilometry (FTP) algorithm. This image pre-filtering stage improves the noise performance of the FTP algorithm and enables the FTP method to demodulate fringe patterns with larger bandwidths. Also, the 2D-CWT technique reduces speckle noise significantly. Moreover, only a single fringe pattern is required in this technique. The 2D-CWT algorithm is capable of separating low frequency terms from the high frequency terms that contain phase-modulated fringe information, even when both interfere, greatly, in the frequency domain. The proposed algorithm is tested, both via computer simulation and using real fringe patterns. This revealed the robustness of this algorithm and also demonstrably enables the demodulation of a wider range of fringe patterns using the FTP technique.     

Single fringe projection profilometry based on sinusoidal intensity normalization and subpixel fitting

A novel fringe projection profilometry using a single sinusoidal fringe pattern projected is proposed. Computer-generated sinusoidal fringe and uniform intensity patterns are firstly projected on a testing object by a liquid crystal display projector. The variable reflection intensity of a fringe pattern is then roughly normalized by division operation applied to the grabbed fringe and uniform intensity patterns projected. Fringe intensity is further normalized by employing an interpolation algorithm. The deformed sinusoidal pattern encoding object shape is converted to a wrapped phase map without using phase-shifting or Fourier transform. Computer simulation and experimental performance are evaluated to demonstrate the validity of the proposed method. The experimental results compared with those of the four-step phase-shifting and fast Fourier transform methods are also presented.     

光载波条纹图的计算机辅助分析——阶梯形虚拟光栅解调法

本文提出一种两维光测条纹图的相位测量技术--阶梯形虚光栅解调算法。此法只需要获取一幅光载波条纹图,由计算机产生两个由N(整数)个象素构成、透射函数呈阶梯形分布、彼此间有一定相位差的两个虚光栅(参考信号),通过在整数域的运算实现对条纹图和虚光栅的相乘型叠加,低通滤波后得到两个相位差已知的叠栅函数,然后从两个相差为已知值的正弦信号中提取相位信息。理论分析和计算机模拟表明,具有相当高的测量精度,计算简便快捷。文中还给出了阶梯虚光栅解调法在投影光栅法三维形貌测量中的应用实例。     

嵌入式三维数字成像系统

提出一种基于数字信号处理器(DSP)的嵌入式三维数字成像系统设计方案。该方案的硬件平台由条纹投影模块、数据采集模块、条纹自动分析模块及储存器等其他辅助电路组成。条纹投影模块将DSP输出的正弦光栅条纹, 经视频编码后用DLP投射到物体表面; 数据采集模块通过CCD相机采集被物体表面三维信息调制后的变形条纹图, 并进行视频解码; 条纹自动分析模块中利用相移算法计算折叠相位, 再结合相位展开算法求绝对相位分布。系统软件采用多线程技术并行控制三个模块。在相位解调过程中以软件流水线为主综合运用了循环展开、数据预取和内联函数等多种方法优化解调程序。实验结果表明, 该系统可以高速、准确地实现三维轮廓测量,优化后相位展开程序速度是优化前的7倍。     

应用神经网络的复杂物体三维测量

将神经网络引入基于结构光投影的复杂物体三维面形测量。在测量过程中,利用神经网络强大的函数逼近能力,得到离散条纹图的连续逼近函数,从中解出物体的相位分布信息,获得物体的三维面形分布。应用神经网络方法,在结构光投影条件下,只需要获取一幅条纹图,便可以完成复杂物体的三维面形测量。该方法相比传统的傅里叶变换轮廓术,不存在滤波操作,不会在测量过程中丢失被测物体的高频分量,具有更高的空间带宽积和灵敏度,能准确测量出复杂物体的细节,更加适用于恢复复杂物体的三维面形。并且该方法在条纹图存在阴影的情况下与傅里叶变换轮廓术相比,能更好地提取出物体的相位信息,恢复物体的三维面形。模拟及实验均验证了该方法的可行性。     

一种针对彩色物体的光栅投影三维测量方法

针对传统彩色编码光栅三维轮廓术中光栅易受到物体表面彩色纹理的干扰,从而造成编码条纹颜色误判和相位误差增大这一问题,提出一种基于互补彩色光栅的三维测量方法,给出了理论分析、光栅设计原理、补偿算法与实验分析。对图像进行初步的解耦校正后,通过预先设计的光栅互补特性,依据彩色响应模型求取物体表面逐点的反射率,并对红绿蓝(RGB)三通道反射率的不平衡进行补偿,消除物体表面彩色纹理的干扰,改善光栅的正弦性。以补偿后的图像来指导彩色编码条纹的分割解码并用傅里叶变换法提取出包裹相位,依据解码结果指导相位展开,继而完成整个三维测量过程。实验证明该方法对彩色纹理的补偿准确有效,降低了彩色纹理对测量的影响。     

Camera calibration with planar crossed fringe patterns

We introduced a novel crossed fringe pattern as the model plane for camera calibration. It can provide abundant control points, in extreme case each pixel on the CCD sensor can be taken as control points. Fourier Fringe Analysis is used to extract the phase distributions from the image of this pattern. Each control point consists of the locations of a pixel and the world coordinates, which can be calculated from the phase distributions. It avoids the complex extracting procedure of those classical control points and could produce abundant control points. Both computer simulation and real data have confirmed the proposed technique is easy to use, reliable and high accurate.     

A two-dimensional fast Fourier transform method for measuring the inclination angle of parallel fringe patterns

In this paper we describe a technique based on two-dimensional fast Fourier fringe pattern analysis for the automated evaluation of the inclination angle of parallel fringe patterns. Analysis of noise-free theoretical parallel fringe patterns is presented and numerical results of the angle evaluation error using this algorithm are discussed. We also demonstrate an application of the method to the measurement of the inclination angle of experimental parallel fringe patterns.     

Comparison of 2D S-Transform Profilometry and 2D Windowed Fourier Transform Profilometry

Both two-dimensional (2D) Windowed Fourier Transform and 2D S Transform are popular time–frequency analysis tools for processing signals. But fewer articles reported the application of 2D S Transform in Fringe pattern analysis. This paper studies the 2D S Transform and compares it with 2D Windowed Fourier Transform. The phase calculation formula based on the 2D Windowed Fourier Transform “ridge” coefficients and 2D S Transform “ridge” coefficients are deduced respectively. A comparison between them is carried out to show their own characteristic in fringe patterns analysis based on the structured-light illumination. It verifies that the 3D reconstruction result of 2D ST is better than that of 2D WFT because of the employment of the adjusted windows by frequency parameters.     

A fast and accurate gamma correction based on Fourier spectrum analysis for digital fringe projection profilometry

In digital fringe projection profilometry, the gamma effect of the used electronic devices seriously affects the accuracy of three-dimensional (3D) topography measurements. Previous gamma correction methods do not take full advantage of the Fourier spectrum of the captured spatial-carrier phase-shifting fringe patterns. Hence, dozens of phase-shifting fringe patterns are required to carry out gamma pre-calibration. In this paper, a fast and accurate gamma correction technique based on a Fourier spectrum analysis is proposed. Only two spatial-carrier fringe patterns with different pre-encoded gamma values are needed and the number of fringe patterns required for gamma pre-calibration is significantly reduced without loss of accuracy. The proposed method is validated by experiments.     

基于自适应条纹投影的高反光物体三维面形测量

结构光投影方法在三维形貌测量中应用广泛,但是由于被测物体表面反射率变化范围较大,过度曝光会导致相位信息无法获取。而传统的高动态范围扫描技术步骤复杂,耗时较长。文中提出一种自适应条纹投影技术,向待测物体表面投射较高灰度级的条纹图,判断并标记过度曝光点。降低投射强度后通过非线性最小二乘法拟合来确定每个饱和像素点最适合的最大输入灰度,用重新生成的自适应条纹图来采集图像并进行相位计算和三维形貌恢复。通过实验验证,该方法可以对物体表面的高反光区域进行有效测量,避免过度饱和,仿真误差在0.02 mm范围内,实测误差约为0.14 mm,实际实验对过曝点的补偿率可达到99%。     

A numerical spatial carrier for single fringe pattern analysis algorithm

Hilbert transform and continuous wavelet transform are combined to form an algorithm for the analysis of a single fringe pattern with open or closed fringes. Only one recorded fringe pattern is needed to extract the phase information. A second π/2 phase shifted fringe pattern should first be generated from the other using Hilbert transform to superpose numerically the spatial carrier. Several phase extraction techniques are conceived to demodulate fringe patterns. In this paper, the wavelet technique is used to obtain the required phase distribution. The algorithm offers an advantage that it is appropriate when the spatial carrier is impossible to be added experimentally. Finally, a simulation was carried out to validate the algorithm, giving good results.     

Phase retrieval of speckle fringe pattern with carriers using 2D wavelet transform

A novel two-dimensional (2D) Gabor continuous wavelet transform (CWT) for phase retrieval and fringe filtering of speckle fringe patterns with spatial carriers is proposed. Theoretical analysis of 2D Gabor CWT is presented and results are compared with advanced fan 2D CWT using both the computer simulated and experimental speckle fringe patterns. It is shown that noise reduction by 2D Gabor CWT demonstrates better results than that of the advanced fan 2D CWT. Two-dimensional Gabor CWT is also compared with 2D Fourier transform and results show that 2D Gabor CWT algorithm has better noise immunity.     

Photorefractive Fourier transform profilometer for the measurement of 3-D object shapes

In the present work dynamic moiré-like fringe patterns, produced by photorefraction with low spatial frequencies, applied for profile determination of small objects is proposed. Basically, a Fourier transform profilometry technique is developed for an automated profile determination. This means, as far as we know, a new experimental procedure that exploits the real time holographic two wave mixing in Bi₁₂TiO₂₀ crystal sample. Besides, the mainly advantages of this procedure are, comparatively to the classical fringe projection method using the Michelson interferometer, best fringe pattern contrast, less speckle noise, absence of noises produced by spurious reflections and, the most significant, a pure sinusoidal shape obtained by dynamic holographic moiré-like process using photorefraction.     

Lens distortion elimination for improving measurement accuracy of fringe projection profilometry

Fringe projection profilometry (FPP) is a powerful method for three-dimensional (3D) shape measurement. However, the measurement accuracy of the existing FPP is often hindered by the distortion of the lens used in FPP. In this paper, a simple and efficient method is presented to overcome this problem. First, the FPP system is calibrated as a stereovision system. Then, the camera lens distortion is eliminated by correcting the captured images. For the projector lens distortion, distorted fringe patterns are generated according to the lens distortion model. With these distorted fringe patterns, the projector can project undistorted fringe patterns, which means that the projector lens distortion is eliminated. Experimental results show that the proposed method can successfully eliminate the lens distortions of FPP and therefore improves its measurement accuracy.     

Noisy fringe pattern demodulation by an iterative phase locked loop

The phase locked loop (PLL) technique applied to demodulate two-dimensional carrier-frequency fringe patterns has been developed recently. Here we present an extension to the basic PLL scheme to demodulate noisy fringe patterns. This modified technique estimates the phase in the fringe pattern iteratively; that is, the first wavefront estimation is done using a flat reference phase and the second iteration takes the demodulated phase found in the first iteration as the new reference. The third demodulating iteration uses the second phase estimation as the reference and so on, until further changes in the detected wavefront fall below a predefined threshold. During the iterative process the bandwidth of the iterative PLL system is gradually decreased to improve the signal-to-noise ratio of the detected phase as well as to resolve noise-generated phase inconsistencies.     

利用强度调制消除零频的傅里叶变换轮廓测量

采用双色正弦条纹投影的傅里叶变换轮廓术中零频分量的扩展对有用基频分量有影响,为此提出一种新型的基于强度调制的零频消除方法提高测量准确度和范围.通过采集一帧变形彩色条纹图并分离颜色通道获得两幅相位差为π的条纹图,对变形条纹做强度调制校正,进而消除零频分量,实现高度解析.理论模拟和实验测试均验证了此方法的有效性.相比传统的相移消除零频分量的方法,此方法具有显著的速度优势,适合于三维物体的实时测量.     

Window fringe pattern demodulation by multi-functional fitting using a genetic algorithm

We present a new way to demodulate complicated fringe patterns containing closed fringes using a genetic algorithm (GA). The entire fringe pattern is divided into a set of partially overlapping smaller sub-image windows. Each of these has a lower dimensionality and as a consequence is faster and can be demodulated more reliably. The demodulation process proceeds row-by-row way passing from one sub-image in a neighborhood until the whole interferogram is processed. The modulating phase of each sub-image is modeled as a parametric analytic function whose parameters are optimized using a GA. The technique is demonstrated demodulating some normalized fringe patterns that have two main difficulties: closed fringes within the interferogram and regions of under-sampled fringes. These fringe images cannot be demodulated by techniques such as the regularized phase tracker (RPT).     

Surface profile measurement of low-frequency vibrating objects using temporal analysis of fringe pattern

A simple and accurate algorithm (phase scanning method) is proposed for 3D surface contouring and dynamic response determination of a vibrating object. A sinusoidal fringe pattern is projected onto a low-frequency vibrating object by a programmable liquid crystal display projector. The fringe patterns are captured by a high-speed CCD camera with a telecentric gauging lens. Phase values are evaluated point by point using phase scanning method. From the phase values of each point on the object, the contour of the specimen at different instants of vibration can be retrieved. In this paper, a small vibrating coin is used to demonstrate the validity of the method and the experimental results are compared with test results on a stationary coin using four-step phase shifting and fast Fourier transform methods. The technique is especially useful in applications where the vibrating object has a complicated shape.     

Automatic window size selection in Windowed Fourier Transform for 3D reconstruction using adapted mother wavelets

Fringe pattern analysis in coded structured light constitutes an active field of research. Techniques based on first projecting a sinusoidal pattern and then recovering the phase deviation permit the computation of the phase map and its corresponding depth map, leading to a dense acquisition of the measuring object. Among these techniques, the ones based on time-frequency analysis permit to extract the depth map from a single image, thus having potential applications measuring moving objects. The main techniques are Fourier Transform (FT), Windowed Fourier Transform (WFT) and Wavelet Transform (WT). This paper first analyzes the pros and cons of these three techniques, then a new algorithm for the automatic selection of the window size in WFT is proposed. This algorithm is compared to the traditional WT using adapted mother wavelet signals both with simulated and real objects, showing the performance results for quantitative and qualitative evaluations of the new method.