[go: up one dir, main page]

WO2006000123A1 - Systeme et procede pour la reconstruction 3d de surface de corps lamellaire et flexible - Google Patents

Systeme et procede pour la reconstruction 3d de surface de corps lamellaire et flexible Download PDF

Info

Publication number
WO2006000123A1
WO2006000123A1 PCT/CN2004/000681 CN2004000681W WO2006000123A1 WO 2006000123 A1 WO2006000123 A1 WO 2006000123A1 CN 2004000681 W CN2004000681 W CN 2004000681W WO 2006000123 A1 WO2006000123 A1 WO 2006000123A1
Authority
WO
WIPO (PCT)
Prior art keywords
sample
image
flexible body
sheet
body surface
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2004/000681
Other languages
English (en)
French (fr)
Inventor
Jinlian Hu
Binjie Xin
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hong Kong Polytechnic University HKPU
Original Assignee
Hong Kong Polytechnic University HKPU
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hong Kong Polytechnic University HKPU filed Critical Hong Kong Polytechnic University HKPU
Priority to PCT/CN2004/000681 priority Critical patent/WO2006000123A1/zh
Publication of WO2006000123A1 publication Critical patent/WO2006000123A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01BMEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
    • G01B11/00Measuring arrangements characterised by the use of optical techniques
    • G01B11/24Measuring arrangements characterised by the use of optical techniques for measuring contours or curvatures
    • G01B11/25Measuring arrangements characterised by the use of optical techniques for measuring contours or curvatures by projecting a pattern, e.g. one or more lines, moiré fringes on the object
    • G01B11/2545Measuring arrangements characterised by the use of optical techniques for measuring contours or curvatures by projecting a pattern, e.g. one or more lines, moiré fringes on the object with one projection direction and several detection directions, e.g. stereo
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/50Depth or shape recovery
    • G06T7/55Depth or shape recovery from multiple images
    • G06T7/564Depth or shape recovery from multiple images from contours
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84Systems specially adapted for particular applications
    • G01N21/88Investigating the presence of flaws or contamination
    • G01N21/89Investigating the presence of flaws or contamination in moving material, e.g. running paper or textiles

Definitions

  • the invention relates to a system and a method for three-dimensional reconstruction of an object surface, which can be widely applied to appearance detection, feature analysis and quality control of an object surface, and is particularly suitable for three-dimensional reconstruction of a sheet-type flexible body surface.
  • Three-dimensional reconstruction technology is widely used in various fields such as industrial production, industrial inspection, space exploration, and scientific research.
  • the required measurement system can be divided into two categories: mechanical contact measurement and photoelectric non-contact measurement.
  • Photoelectric non-contact measurement can be further divided into: binocular or multi-eye stereo measurement, structural grating, laser triangulation, laser flying spot measurement, Moir interferometry and other methods.
  • Each type of 3D reconstruction technology has its own advantages and disadvantages.
  • Binocular or multi-view stereo measurement in photoelectric non-contact measurement is based on binocular stereo matching algorithm and camera calibration technology. For surface surfaces without many texture details or dark colors, the measurement accuracy is poor and unstable.
  • Structural grating measurement is based on the principle of coded grating and triangulation. For the surface of a mottled or multi-patterned object, the reconstruction algorithm will fail due to the destruction of the coding pattern of the grating.
  • Laser triangulation and laser flying spot measurement are a point-by-point measurement method with low scanning efficiency and high cost.
  • Mohr interferometry The amount is also not suitable for measuring the surface of mottled, multi-pattern or dark black objects, and the automation of Moiré analysis technology for complex surfaces is also a technical difficulty.
  • the surface analysis of a sheet-type flexible body material is mainly based on a two-dimensional surface image obtained by light reflection or projection, and the two-dimensional surface image cannot express three-dimensional information and features of the surface of the material.
  • appropriate 3D reconstruction techniques are needed to extract 3D information and features from the surface of the material. Due to the complexity of the surface of the material, variegated, multi-patterned, dark or black, lack of texture details, etc. often occur. In the face of such complex sample testing types, the selection of several reconstruction techniques described above can only be applied to specific Sample type, cannot cover most sample types. There is therefore an urgent need to provide a three-dimensional reconstruction technique for a versatile flexible material surface that can be applied to various types of flexible material surfaces. Summary of the invention
  • an object of the present invention is to provide a technique for three-dimensional reconstruction which is not affected by factors such as color, texture and gloss of the surface of the sample, and which can quickly and accurately obtain three-dimensional information of the surface of the flexible material.
  • the present invention provides a system for three-dimensional reconstruction of a sheet-type flexible body surface, comprising: a background light source for emitting uniform diffuse reflected light having a certain brightness; and a sample holding device Holding a sample to be three-dimensionally reconstructed thereon, the sample holding device has a vertex, the sample is bent at the vertex, and the diffuse light emitted from the background light source is projected to the test
  • An image collecting device for collecting an image of the sample to obtain a side projection image of the sample at the vertex; a sample driving device driven to be clamped in the test
  • the sample on the gripping device passes uniformly through the apex, so that the image is taken
  • the collecting device can sequentially obtain the side projection images at different positions on the surface of the sample; a computer controlling the movement of the sample driving device and receiving the sequence of the side projection image of the sample collected by the image collecting device, The image sequence is processed and analyzed, and the side projection height information of each frame image is extracted, and a three-dimensional ster
  • a method for three-dimensional reconstruction of a sheet-type flexible body surface comprising the steps of:
  • Sampling collecting a side projection image at each position of a sample to form a side projection image sequence
  • Contour extraction processing each of the side projection images to extract side projection height information of each frame image
  • the obtained side projection height information is spliced according to the acquisition order, thereby obtaining a three-dimensional stereoscopic image of the surface of the sample.
  • the invention has the beneficial effects that, because the invention is based on the characteristic that the sheet-type flexible body is easy to bend, the sheet-shaped material is fixed and bent by a curved type sample holding device to obtain a strip-shaped convex surface with a certain curvature; Under the illumination of a backward diffuse reflection light source with a certain brightness, an image acquisition device is used to acquire an image to obtain a side projection image of a strip-shaped convex surface; in addition, the sample surface is continuously rotated by a computer-controlled sample driving device to obtain a sample surface.
  • the present invention is a contactless photoelectric measuring method specifically adapted for three-dimensional reconstruction of a sheet-type flexible body surface, which is not subject to the color of the surface of the sample.
  • the three-dimensional information of the surface of the flexible material can be obtained quickly and accurately by the influence of color, texture and gloss. It is suitable for offline detection and online inspection. It can be widely used in textile, printing, packaging and other industries.
  • FIG. 1 is a schematic structural view of a three-dimensional reconstruction system for a sheet-type flexible body surface of the present invention
  • FIGS. 2A and 2B show two embodiments of a curved-type sample holding device of the present invention
  • FIGS. 3A and 3B The manner of the backward illumination in the present invention is respectively shown;
  • Figure 4 is a side projection image of the sample after bending
  • Figure 5 is a side projection height curve of the surface of the sample
  • Fig. 6 is a view showing the configuration of a computer in the present invention.
  • the surface three-dimensional reconstruction system of the present invention is mainly applicable to a sheet-type flexible body material.
  • a flexible body material is defined as: A non-rigid solid material that bends with a certain curvature under a small external force.
  • the sheet-type material is defined as: a continuum having a certain thickness and area size, a large area size which is many times larger than the thickness (usually 20 times or more), and a small difference in thickness.
  • a system for three-dimensional reconstruction of a sheet-type flexible body surface of the present invention comprises: a background light source 2, a sample holding device 3, a sample driving device 5, an image pickup device 7, and a computer 10.
  • the background light source 2 is an array of point sources or point sources that emit divergent beams for emitting a uniform diffusely reflected light having a uniform brightness.
  • the background light source 2 can also be equipped with light A system (not shown) that causes the divergent beam to emit a parallel beam or a concentrated beam after passing through the optical system.
  • the background light source may be a visible light source, or may be an infrared light source or an ultraviolet light source.
  • the sample holding device 3 is curved, and a sample 4 to be three-dimensionally reconstructed is fixed on the sample holding device 3.
  • the sample holding device has a vertex 33 at which the sample 4 is bent.
  • the diffused light emitted from the background light source 2 is projected onto the sample 4.
  • the image capture device 7 is disposed on one side of the sample 4, which may be a CCD camera, a CMOS camera, a digital camera or a digital video camera, which is connected to the computer 10, so that the computer 10 can collect the image in real time through the image capture device 7.
  • the image of the sample 4 gives a sequence of side projection images of the sample 4 at the apex 33.
  • the sample driving device is a motor 5, which is preferably a stepping motor, which is connected to the computer 10, and is driven and controlled by the computer 10.
  • the motor 5 drives the sample 4 clamped on the sample holding device 3 to uniformly pass through the apex 33.
  • the sample 4 is bent at a vertex 33 at a certain position in the longitudinal direction.
  • each position of the longitudinal direction of the sample 4 can be bent at the apex 33, so that the image acquisition device 7 can sequentially obtain different surface of the sample.
  • a side projection image at the location the motor 5 can directly drive the sample holding device 3, and the sample holding device can also be driven by an intermediate transmission mechanism (not shown), which is generally a reduction gear set, so that The sample is moved at an appropriate speed.
  • the computer 10 of the system is a control center of the entire system and a processing center of the image, as shown in FIG. 6, which includes an image capture card 101, a motor driver 102, and a conventional central processing unit CPU and memory.
  • the image capture card and The image acquisition device 7 is connected for digitizing the acquired image and storing it in a memory.
  • the motor driver is coupled to the motor 5 for controlling the motor 5 to operate at a predetermined speed or step.
  • the image processing module 104 further includes a contour extraction module 1041 and a splicing module 1042, the contour extraction module 1041 is configured to extract side projection height information of each frame image, and the splicing module 1042 is configured to project the side projection The height information is spliced in the sampling order to obtain a three-dimensional image of the surface of the sample.
  • the contour extraction module and the splicing module described above may be implemented by an electronic circuit or a software program running in the computer 10.
  • the sample holding device has a curved shape for the flexible material of the present invention. It may be a spike-shaped "A" type as shown in Fig. 2A or a circular-arc "0" type as shown in Fig. 2B, and may include any form formed by planar bending. A common feature of this curved form is that it has a vertex 33 so that each portion of the sample 4 can project a contour onto the image capture device 7 as it passes through the apex 33, completing the data sampling. As shown in Fig.
  • the holder 31 includes three rollers 311 arranged in an "A" shape and a conveyor belt surrounding the roller. 312.
  • the three rollers 311 are rotated by the motor 5 so that the belt 312 is rotated around the rollers, so that the sample 4 fixed to the belt 312 by the chuck 32 also moves with the belt.
  • Fig. 2B shows another embodiment of the sample holding device, which is also composed of a holder 31' and a collet 32.
  • the holder 31' is a roller having a larger diameter
  • the motor 5 drives the roller to rotate
  • the sample 4 fixed to the roller by the chuck 32 rotates with the roller.
  • the collet 32 is also available in a variety of embodiments, either as a collet chuck or as a magnetic collet.
  • the device has a detailed description in the utility model patent application filed by the inventor of the present application (application number: 03207803. x, invention name: magnetic cloth sample clamping device for digital fabric appearance/surface test assessor), No longer.
  • the background light source 2, the sample holding device 3 and the image pickup device 7 are configured in a manner of rearward illumination to obtain a projected image of the highest contrast.
  • the rear illumination mode is realized by arranging the sample 4 between the background light source 2 and the image pickup device 7.
  • the background light source 2 is illuminated from the rear of the sample 4, and a sharp outline of the sample is projected at the image capture device 7.
  • the backward illumination mode can also be implemented in another manner.
  • the background light source 2 and the image acquisition device 7 are located on the same side of the sample 4, and a back reflection is disposed behind the sample 4.
  • the mirror 80 is provided with a beam splitter 81 between the background light source 2 and the image acquisition device 7, and the back illumination is realized by back reflection.
  • a sealed space 1 capable of isolating natural light, and at least the sample 4, the background light source 2, and the image capturing device 7 are disposed in the sealed space 1, so that a clearer projection can be obtained.
  • the confined space 1 can be implemented as a ruled geometry or as an irregular geometry.
  • Fig. 4 shows a side projection image of a sample at a position of the apex 33, which clearly reflects its outline.
  • the image processing module of the present system extracts the side projection height information of the sample surface of each frame image in the image sequence as shown in FIG.
  • the present invention also provides a method for three-dimensional reconstruction of a sheet-type flexible body surface, comprising the following steps: Image sampling: collecting a side projection image at each position of a sample to form a side projection image sequence; contour extraction: processing the side projection images to extract side projection height information of each frame image; stitching: according to the collection The obtained side projection height information is sequentially spliced to obtain a three-dimensional stereoscopic image of the surface of the sample.
  • System startup The computer, motor, light source, and photoelectric camera system are in working state
  • System commissioning calibration Check whether all parts of the system are working properly.
  • the sampling step further includes the step of: clamping the sample 4 on the sample holding device 3 such that the sample 4 is bent at a vertex 33 of the sample holding device 3;
  • a background light source 2 is disposed on one side of the sample 4; the sample is driven to be continuously bent through the apex 33 of the sample holding device 3; and the side projection of the sample 4 at the vertex position is collected by the image acquisition device 7.
  • An image; the side projection image is projected by the background light source 2.
  • the algorithm in the three-dimensional reconstruction method and system for the sheet-type flexible body surface mainly focuses on how to separate the side projection height information from the side projection image, wherein the algorithms applicable to the step include:
  • the histogram is defined as the grayscale distribution of the image.
  • the grayscale image of the side projection of the curved convex surface of the specimen is f (x, y), where ( ⁇ , y) is the image coordinate, and f (x, y) is the gray of the pixel at ( ⁇ , y) Degree, f (x, y) G [0, 255] o where a gray value of 0 indicates that the pixel is white, and a gray value of 255 indicates that the pixel is black.
  • Its histogram is h ( i ), ie [0, 255].
  • Image segmentation Set the threshold t u u, +u 2 + ⁇ ⁇ ⁇ 2 according to the gray scale distribution of the background and the sample
  • Edge detection is performed by edge detection algorithm. Common edge detection algorithms include Marr edge detection method, Sobel operator, Robert operator, Laplacian operator and so on.
  • the Robert operator is an operator that uses local difference operators to find edges. It is given by: Where / is an input image with integer pixel coordinates.
  • the algorithm for edge detection described above is discussed in detail in the reference “Digital Image Processing, Prentice Hall, 1998" and will not be repeated here.
  • Method 3 Radon Transform: The gray value is summed according to the vertical projection of the image (that is, the Norton transform in the vertical direction), and the gray scale distribution of the image in the vertical direction is obtained. Since the accumulation of gradation is linear with the thickness of the sample, the surface profile height of the sample can be obtained by dividing by a certain proportional coefficient.
  • the gray value of the projected image at the (x, y) point is /( ⁇ , where xe [l, M] is the image abscissa, and ye [l, N] is the image ordinate, then the frame
  • the method for three-dimensional reconstruction of a sheet-type flexible body surface of the present invention can be used for both off-line detection and on-line detection.
  • the flexible body is easy to bend, and the thin-type sample holding device is used to fix and bend the sheet-shaped material to obtain a strip-shaped convex surface with a certain curvature; the sample is contained in a sealed space capable of isolating natural light, and has uniform brightness.
  • the image is acquired by the photoelectric imaging system to obtain a side projection image of the strip-shaped convex surface; the sample is driven continuously by the computer-driven sample driving device, and the surface of the sample is obtained at different positions.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Length Measuring Devices By Optical Means (AREA)

Description

薄片型柔性体表面三维重建的系统和方法 技术领域
本发明涉及一种对物体表面进行三维重建的系统和方法,可广泛应用 于物体表面的外观检测、特征分析和质量控制, 尤其适用于薄片型柔性体 表面的三维重建。 背景技术
三维重建技术被广泛地应用于工业生产、 工业检测、 宇宙探索、 科学 研究等各个领域。一般意义上讲, 为获得物体的三维形状, 所需的测量系 统可以分为: 机械接触测量和光电式非接触测量两大类。光电式非接触测 量又可以分为: 双目或多目立体测量, 结构光栅, 激光三角测量, 激光飞 点测量, 莫尔干涉测量等方法。每一种三维重建技术都有各自的优点和缺 点。
机械接触测量主要适用刚体的连续表面, 具有较高精度和稳定性。而 对于所谓的薄片型柔性体, 其包括: 纺织品, 塑料薄膜, 纸张和其他可自 由弯曲的薄片型材料, 则无法采用该机械接触测量。光电式非接触测量中 的双目或多目立体测量, 是基于双目立体匹配算法和相机标定技术, 对于 没有很多纹理细节或者深色的物体表面, 测量精度差且不稳定。 结构光栅 测量,是基于编码光栅和三角测量的原理,对于杂色或多花紋的物体表面, 由于破坏了光栅的编码图案, 会导致重建算法失败。激光三角测量和激光 飞点测量, 是一种逐点测量方法, 扫描效率低, 且成本昂贵。 莫尔干涉测 量, 同样不适合测量杂色, 多花纹或深黑色的物体表面, 而且复杂曲面的 莫尔紋分析技术的自动化也是技术的难点。
在工业生产中,对于薄片型柔性体材料的表面分析主要是基于光反射 或投射得到的二维表面图像,而二维表面图像无法表达材料表面的三维信 息和特征。 在此背景下, 需要合适的三维重建技术来提取材料表面的三维 信息和特征。 由于材料表面的复杂性, 杂色、 多花纹、 深色或黑色, 缺少 紋理细节等情况经常出现, 面对如此复杂的样本测试种类, 选用前面所述 的几种重建技术只能适用于特定的样本类型, 无法覆盖大部分样本类型。 因此迫切需要提供一种通用的柔性材料表面的三维重建技术,其能够适用 于各种类型的柔性材料表面。 发明内容
鉴于上述问题, 本发明的目的是提供一种三维重建的技术, 其不受试 样表面的色彩, 纹理和光泽等因素的影响, 可快速、 准确地得到柔性材料 表面的三维立体信息。
为了实现上述目的,本发明提供了一种用于薄片型柔性体表面三维重 建的系统, 包括: 一背景光源, 用于发出均匀且具有一定亮度的后向漫反 射光; 一试样夹持装置, 其上夹持有待三维重建的一试样, 所述试样夹持 装置具有一顶点, 所述试样在该顶点处弯折, 从所述背景光源发出的漫反 射光投射到该试样上; 一图像釆集装置, 用于釆集所述试样的图像, 从而 得到所述试样在所述顶点处的侧面投影图像; 一试样驱动装置, 其驱动夹 持在所述试样夹持装置上的所述试样均匀地通过所述顶点,从而该图像采 集装置能够依次得到该试样表面不同位置处的侧面投影图像; 一计算机, 其控制该试样驱动装置动作, 同时接收该图像采集装置采集到的所述试样 的侧面投影图像的序列, 对所述图像序列进行处理和分析, 提取每一帧图 像的侧面投影高度信息, 并按照采集顺序拼接得到所述试样表面的三维立 体图像。
本发明的另一方面,还提供了一种用于薄片型柔性体表面三维重建的 方法, 包括如下步骤:
采样: 采集一试样的各个位置处的侧面投影图像, 形成侧面投影图像 序列;
轮廓提取: 对所述各个侧面投影图像进行处理, 提取每一帧图像的侧 面投影高度信息;
拼接: 按照采集顺序将得到的侧面投影高度信息进行拼接, 从而得到 所述试样表面的三维立体图像。
本发明的有益效果是, 由于本发明基于薄片型柔性体易于弯曲的特 点, 利用一种曲面型试样夹持装置, 将薄片型材料固定并弯曲获得一定曲 率的条形凸面; 通过均勾且具有一定亮度的后向漫反射光源照射下, 利用 图像采集装置来采集图像, 得到条形凸面的侧面投影图像; 另外, 通过由 计算机控制的试样驱动装置, 连续转动试样, 得到试样表面在不同位置连 续弯曲形成的条形凸面的侧面投影图像系列; 利用图像处理和分析技术, 提取图像系列中每一帧图像的试样表面的侧面投影高度信息,并按照采集 顺序拼接提到试样表面的三维立体图像。所以本发明是一种专门适用于薄 片型柔性体表面的三维重建的无接触光电测量方法,其不受试样表面的色 彩, 纹理和光泽等因素的影响, 可快速、 准确地得到柔性材料表面的三维 立体信息。 适用于离线检测和在线检测两种模式, 可广泛应用于纺织, 印 刷, 包装等行业。
下面结合附图和具体实施方式对本发明作进一步说明。 附图说明
图 1是本发明用于薄片型柔性体表面的三维重建系统的结构示意图; 图 2A和图 2B示出了本发明中的曲面型试样夹持装置的两种实施例; 图 3A和图 3B分别示出本发明中的后向照明的方式;
图 4是试样弯曲后的侧面投影图像;
图 5是试样表面的侧面投影高度曲线;
图 6是本发明中的计算机的构成示意图。
其中, 附图标记说明如下:
光密闭空间 1
背景光源 2
试样夹持装置 3
支架 31, 31'
辊子 311
传送带 312
夹头 32
顶点 33
试样 4 图像采集装置 7
后反射镜 80
分束镜 81
计算机 10
图像采集卡 101
电机驱动器 102
图像处理模块 104
轮廓提取模块 1041
拼接模块 1042 具体实施方式
以下结合附图和实例对本发明作进一步的描述。
本发明的表面三维重建的系统主要适用于薄片型柔性体材料。柔性体 材料的定义为: 非刚性固体材料, 其在较小的外力下能进行一定曲率的弯 曲。 薄片型材料的定义为: 具有一定的厚度和面积尺寸, 面积尺寸相对厚 度大很多倍 (通常在 20倍以上), 且厚度变化差异不大的连续体。
参照图 1所示, 本发明用于薄片型柔性体表面三维重建的系统包括: 背景光源 2, 试样夹持装置 3,试样驱动装置 5, 图像采集装置 7和计算机 10。
所述背景光源 2是发出发散光束的点光源或点光源的阵列,其用于发 出均匀且具有一定亮度的后向漫反射光。 另外, 该背景光源 2还可配有光 学系统 (未示出), 使得所述发散光束经该光学系统后发出平行光束或会 聚光束。 而且, 该背景光源可以是可见光源, 也可以是红外光光源或者紫 外光光源。
该试样夹持装置 3呈曲面型,待三维重建的一试样 4固定在该试样夹 持装置 3上, 所述试样夹持装置具有一顶点 33, 试样 4在该顶点处弯折, 从背景光源 2发出的漫反射光投射到该试样 4上。
该图像采集装置 7设置在该试样 4的一侧, 其可为 CCD摄像头、 CMOS 摄像头、 数码相机或数码摄像机, 其连接到计算机 10上, 从而计算机 10 可通过该图像采集装置 7实时地采集该试样 4的图像,得到试样 4在顶点 33处的侧面投影图像序列。
该试样驱动装置在本实施例中为一电机 5, 其优选为步进电机, 该电 机 5连接到计算机 10上, 由该计算机 10进行驱动控制。 该电机 5驱动夹 持在试样夹持装置 3上的试样 4均匀地通过所述顶点 33,由图 1可以看出, 试样 4沿纵向的某一位置会在顶点 33处发生弯折, 这样, 随着该试样匀 速地被驱动通过该顶点, 该试样 4纵向的每个位置均可在该顶点 33处发 生弯折,从而该图像采集装置 7能够依次得到该试样表面不同位置处的侧 面投影图像。 另外, 该电机 5可以直接驱动该试样夹持装置 3, 也可通过 一中间传动机构(未示出)驱动所述试样夹持装置, 该中间传动机构一般 为一减速齿轮组, 从而可以适当的速度驱动该试样运动。
本系统的计算机 10是整个系统的控制中心和图像的处理中心,如图 6 所示, 其除了包括常规的中央处理单元 CPU和存储器之外, 还包括一图像 采集卡 101、 一电机驱动器 102和一图像处理模块 104。 该图像采集卡与 所述图像采集装置 7相连, 用于将采集到的图像数字化, 并将其存储到存 储器中。 所述电机驱动器与所述电机 5相连, 用于控制所述电机 5以预定 的速度或步长运转。
该图像处理模块 104还进一步包括一轮廓提取模块 1041和一拼接模 块 1042, 所述轮廓提取模块 1041用于提取每一帧图像的侧面投影高度信 息, 所述拼接模块 1042用于将所述侧面投影高度信息按采样顺序拼接得 到所述试样表面的三维立体图像。上述轮廓提取模块和拼接模块可以由电 子线路实现, 也可是在计算机 10中运行的软件程序。
如图 2A和 2B所示, 该试样夹持装置针对本发明的柔性材料, 具有弯 曲的形态。 其可以是图 2A所示的尖峰形的 "A"型, 也可以是图 2B所示 的圆弧形的 "0"型, 而且还可以包括平面弯曲所形成的任何形态。 该弯 曲形态的共同特征是具有一顶点 33,这样,才能够使得试样 4的每一部分 能够在通过该顶点 33时, 将轮廓投影到图像采集装置 7上, 完成数据采 样。 如图 2A所示, 是该试样夹持装置的第一种实施例, 其由支架 31和夹 头 32组成, 支架 31包括布置为 "A"字形的三个辊子 311和环绕该辊子 的传送带 312。 该三个辊子 311在电机 5的驱动下转动, 从而使得该传送 带 312环绕该辊子转动, 因此, 由夹头 32固定在该传送带 312上的试样 4 也随着该传送带运动。
图 2B 示出了该试样夹持装置的另一实施例, 其也由支架 31 '和夹头 32组成。 所 同的是该支架 31 '是一直径较大的辊子, 电机 5驱动该辊子 转动, 由夹头 32固定在该辊子上的试样 4随该辊子一起转动。 该夹头 32 也有多种实施方式, 可以是弹簧夹头, 也可是磁性夹头。 有关该试样夹持 装置, 在本申请发明人提交的实用新型专利申请 (申请号: 03207803. x, 发明名称: 用于数码织物外观 /表面测试评定仪的磁性布样夹持装置) 中 有详细的描述, 在此不再赘述。
在本发明的系统中, 该背景光源 2、 试样夹持装置 3和图像采集装置 7配置为后向照明的方式, 以获得最高对比度的投影图像。 如图 3A所示, 该后向照明方式通过将该试样 4设置在该背景光源 2和该图像采集装置 7 之间来实现。 这样, 该背景光源 2从该试样 4的后方照明, 就会在该图像 采集装置 7处投射出该试样的清晰的轮廓。
另外, 该后向照明方式也可以另一种方式实现, 如图 3B所示, 该背 景光源 2与图像采集装置 7位于试样 4的同一侧,在所述试样 4的后方设 置有一后反射镜 80,在所述背景光源 2和所述图像采集装置 7之间设置一 分束镜 81, 通过后反射来实现该后向照明。
为了实现最佳的图像采集效果,最好提供一个可以隔绝自然光的密闭 空间 1,至少将试样 4、背景光源 2和图像采集装置 7设置在该密闭空间 1 中, 从而可以得到更加清晰的投影图像。密闭空间 1的实现方式可以是规 则的几何体, 也可以是不规则的几何体。
图 4示出了一试样处于顶点 33的某一位置处的侧面投影图像, 该图 像清晰地反映了其外形轮廓。
利用图像处理和分析技术,本系统的图像处理模块提取图像序列中的 每一帧图像的试样表面的侧面投影高度信息如图 5所示。
与本发明的柔性体表面三维重建系统相对应,本发明还提供了一种用 于薄片型柔性体表面三维重建的方法, 其包括如下步骤: 图像采样: 采集一试样的各个位置处的侧面投影图像, 形成侧面投影 图像序列; 轮廓提取: 对所述各个侧面投影图像进行处理, 提取每一帧图 像的侧面投影高度信息; 拼接: 按照采集顺序将得到的侧面投影高度信息 进行拼接, 从而得到所述试样表面的三维立体图像。
另外, 在上述步骤之前, 还包括下列的初始化步骤: 系统启动: 使计 算机, 电机, 光源, 光电摄像系统处于工作状态; 系统调试校准: 检查系 统的各部分是否工作正常。
在所述的采样步骤还进一步包含如下步骤- 将试样 4夹持在试样夹持装置 3上,使得所述试样 4在该试样夹持装 置 3的一顶点 33处弯折;在试样 4一侧设置一背景光源 2;驱动所述试样 连续弯折通过所述试样夹持装置 3的该顶点 33;由图像采集装置 7采集位 于该顶点位置的试样 4的侧面投影图像;该侧面投影图像由所述背景光源 2投射形成。
在该薄片型柔性体表面三维重建方法和系统中的算法主要集中在如 何从该侧面投影图像中分离出侧面投影高度信息,其中可用于该步骤的算 法包括:
方法 1 :
•直方图分析: 直方图定义为图像的灰度分布曲线。 令试样弯曲凸面 的侧面投影的灰度图像为 f (x, y), 其中, (χ, y)是图像坐标, f (x, y) 是 (χ, y) 处的象素点的灰度值, f (x, y) G [0, 255] o 其中, 灰度值 为 0代表该象素点为白色, 灰度值为 255代表该象素点为黑色。 它的直方 图为 h ( i ), i e [0, 255]。 取投影图像序列中某一投影图像, 图像中存在 两类区域:背景区域和试样投影区域,它们的灰度分布模型接近高斯分布。 假设两个分布的均值和方差分别为 (ul, δΐ) 和 (U2, 52), 则直方图 的分布模型为:
(Ζ·— Μ2) y{i) = A -e 1 +A 通 过 寻 找 函 数
MSE ∑[ 0- WO]2的极小值得到两个高斯分布的最佳估计值(ul, δΐ)
Ν ζ = 1
和 (u2, δ 2),
图像分割: 根据背景和试样的灰度分布, 设定阈值 t 二 u、 +u2 + λ{δ ~ δ2)
2 λ为经验系数, 然后进行二值化处理, f (x, y)
Figure imgf000012_0001
•高度提取: 经上述二值化处理的图像中, 象素值为 255的区域为试 样, 象素值为 0的区域为背景, 根据这一特征, 扫描图像, 得到试样和背 景的边界轮廓线, 并确定边界轮廓线上每一点的坐标, 其中沿垂直边界方 向的坐标为高度坐标。
方法 2:
•边缘检测: 利用边缘检测的算法进行边缘检测, 常见的边缘检测算 法包括 Marr边缘检测方法, Sobel算子, Robert算子, 拉普拉斯算子等。
Robert 算子是一种利用局部差分算子寻找边缘的算子。 它由下式给 出:
Figure imgf000012_0002
其中, / 是具有整数象素坐标的输入图像。 对于上述边缘检测的 算法, 在参考文献 "数字图像处理, Prentice Hall 著, 1998 " 中有详细 论述, 在此不再赘述。
•高度提取: 扫描图像, 根据检测出来的轮廓边缘, 确定边界轮廓线 上每一点的坐标, 其中沿垂直边界方向的坐标为高度坐标。
方法 3: 诺顿变换 (Radon Transform) : 根据图像在垂直方向的投影 对灰度值进行求和 (即垂直方向上的诺顿变换), 得到图像在垂直方向上 的灰度分布。 由于灰度的累积跟试样的厚度成线性关系, 所以除以一定的 比例系数, 就可以得到试样的表面轮廓高度。
具体地说, 令投影图像在 (x, y) 点的灰度值为 /(χ, , 其中 x e [l, M]为图像横坐标, y e [l, N]为图像纵坐标, 则此帧图像在垂直方向的诺 顿变换 ?(x) = fix, 。即对于水平方向的每一位置 x,将图像灰度值沿垂 y = l
直方向累积求和,得到水平位置 X处的诺顿变换 R (x), 对所有水平位置 X ≡ [1 , M]进行上述诺顿变换, 即提取图像在垂直方向上的灰度分布; 由于 沿垂直方向的该灰度的累积 R G 跟该水平位置处试样的厚度成线性关 系,所以将该累积值 R (x)除以一定的比例系数,即可得到试样在横坐标 (水 平位置) 为 X处的表面轮廓高度。
另外, 本发明的用于薄片型柔性体表面三维重建的方法, 可以用于离 线检测和在线检测两种模式。 工业实用性
本发明的用于薄片型柔性体表面三维重建的系统和方法,基于薄片型 柔性体易于弯曲的特点, 利用曲面型试样夹持装置, 将薄片型材料固定并 弯曲获得一定曲率的条形凸面;在一个可以隔绝自然光的密闭空间内容纳 试样, 在均匀且具有一定亮度的后向漫反射光源照射下, 利用光电摄像系 统采集图像, 得到条形凸面的侧面投影图像; 通过由计算机控制的试样驱 动装置, 连续均勾地转动试样, 得到试样表面在不同位置连续弯曲形成的 条形凸面的侧面投影图像系列; 利用图像处理和分析技术, 提取图像系列 中每一帧图像的试样表面的侧面投影高度信息,并按照采集顺序拼接得到 试样表面的三维立体图像。 由上述可知, 本发明所提出的三维重建系统和 方法不受试样表面的物理性质影响, 无论是有条紋的试样还是深色的试 样, 均可获得该试样表面的精确的三维模型。所以有效地克服了现有技术 的缺陷。
虽然本发明已以上述实施例揭示如上,但是本发明并不局限于该实施 例的内容,本技术领域的普通技术人员,在不脱离本发明的精神和范围内, 可以对本发明做出任意的更动和润饰。本发明的保护范围以所附的权利要 求书为准。

Claims

权利要求
1、 一种用于薄片型柔性体表面三维重建的系统, 其特征在于, 包括: 一背景光源, 用于发出均匀且具有一定亮度的后向漫反射光; 一试样夹持装置, 其上夹持有待三维重建的一试样, 所述试样夹持装 置具有一顶点, 所述试样在该顶点处弯折, 从所述背景光源发出的漫反射 光投射到该试样上;
一图像采集装置, 用于采集所述试样的图像, 从而得到所述试样在所 述顶点处的恻面投影图像;
一试样驱动装置,其驱动夹持在所述试样夹持装置上的所述试样均匀 地通过所述顶点,从而该图像采集装置能够依次得到该试样表面不同位置 处的侧面投影图像;
一计算机, 其控制该试样驱动装置动作, 同时接收该图像采集装置采 集到的所述试样的侧面投影图像的序列,并对所述图像序列进行处理和分 析, 提取每一帧图像的侧面投影高度信息, 按照采集顺序将该每一帧图像 的侧面投影高度信息拼接得到所述试样表面的三维立体图像。
2、 如权利要求 1所述的薄片型柔性体表面三维重建的系统, 其特征 在于, 所述背景光源是发出发散光束的点光源或点光源的阵列, 并可配有 光学系统, 使得所述发散光束经该光学系统后发出平行光束或会聚光束; 所述背景光源可以是可见光源, 也可以是红外光光源或者紫外光光源。
3、 如权利要求 1所述的薄片型柔性体表面三维重建的系统, 其特征 在于, 所述试样夹持装置具有一支架 (31, 31' ) 和多个夹头 (32), 所述 顶点形成在所述支架的顶端, 所述试样由所述夹头夹持在所述支架上, 固 定在该试样夹持装置上的试样以一定的曲率弯曲,便于采集该试样不同位 置处的轮廓。
4、 如权利要求 3所述的薄片型柔性体表面三维重建的系统, 其特征 在于,所述支架(31 )包括布置成" A"形的三条辊子(311 )和传送带 312, 所述传送带环绕所述辊子, 形成所述支架; 所述夹头将该试样夹持在所述 传送带上。
' 5、 如权利要求 3所述的薄片型柔性体表面三维重建的系统, 其特征 在于, 所述支架 (31' ) 是一大直径的辊子。
6、 如权利要求 1所述的薄片型柔性体表面三维重建的系统, 其特征 在于, 所述试样驱动装置是电机 (5)。
7、 如权利要求 1所述的薄片型柔性体表面三维重建的系统, 其特征 在于, 所述图像采集装置是选自下列集合的其中之一: CCD摄像头、 CMOS 摄像头、 数码相机和数码摄像机。
8、 如权利要求 6所述的薄片型柔性体表面三维重建的系统, 其特征 在于, 所述计算机包括一图像采集卡、 一电机驱动器和一图像处理模块
( 104); 所述图像采集卡与所述图像采集装置相连, 用于将该图像采集装 置采集到的图像数字化, 并将其存储在存储器中; 所述电机驱动器与所述 电机相连, 用于控制所述电机的运转。
9、 如权利要求 8所述的薄片型柔性体表面三维重建的系统, 其特征 在于, 所述图像处理模块 (104) 还包括一轮廓提取模块 (1041 ) 和一拼 接模块 (1042), 所述轮廓提取模块用于提取每一帧图像的侧面投影高度 信息,所述拼接模块用于将所述提取到的每一帧图像的侧面投影高度信息 按采样顺序拼接, 从而得到所述试样表面的三维立体图像。
10、 如权利要求 1所述的薄片型柔性体表面三维重建的系统, 其特征 在于, 所述的背景光源、 试样夹持装置和图像采集装置配置为后向照明的 方式, 以获得最高对比度的投影图像。
11、 如权利要求 10所述的薄片型柔性体表面三维重建的系统, 其特 征在于, 所述试样夹持装置设置在所述背景光源和所述图像采集装置之 间, 以实现后向照明。
12、 如权利要求 10所述的薄片型柔性体表面三维重建的系统, 其特 征在于, 所述的背景光源与图像采集装置位于所述试样夹持装置的同一 侧, 一后反射镜 (80) 设置在所述试样的后方, 一分束镜(81 ) 设置在所 述背景光源和所述图像采集装置之间, 通过后反射来实现该后向照明。
13、 如权利要求 6所述的薄片型柔性体表面三维重建的系统, 其特征 在于, 所述电机通过一中间传动机构与所述试样夹持装置连接。
14、 一种用于薄片型柔性体表面三维重建的方法, 其特征在于, 包括 如下步骤:
采样: 采集一试样的各个位置处的侧面投影图像, 形成侧面投影图像 序列;
轮廓提取: 对所述各个侧面投影图像进行处理, 提取每一帧图像的侧 面投影高度信息;
拼接: 按照釆集顺序将得到的侧面投影高度信息进行拼接, 从而得到 所述试样表面的三维立体图像。
15、 如权利要求 14所述的用于薄片型柔性体表面三维重建的方法, 其特征在于, 所述采样步骤还进一步包含如下步骤:
将所述试样夹持在一试样夹持装置上,使得所述试样在该试样夹持装 置的一顶点处弯折;
在所述试样一侧设置一背景光源;
驱动所述试样连续弯折通过所述试样夹持装置的该顶点; .
由一图像采集装置采集位于该顶点位置的所述试样的侧面投影图像; 该侧面投影图像由所述背景光源投射形成。
16、 如权利要求 14所述的用于薄片型柔性体表面三维重建的方法, 其特征在于, 所述的轮廓提取步骤还包括如下三个步骤:
直方图分析: 对试样弯曲凸面的侧面投影图像进行直方图分析, 分别 得出投影图像中背景和试样两个部分的灰度的高斯分布均值和方差; 图像分割, 根据该背景和试样的灰度分布均值和方差, 设定一阈值, 根据该阈值对图像进行二值化处理, 即象素点的灰度值大于该阈值, 则该 象素点的象素值为 255, 象素点的灰度值小于该阈值, 则该象素点的象素 值为 0 ;
高度提取, 根据图像中象素值为 0的区域为背景, 象素值为 255的区 域为试样, 扫描图像, 得到背景与试样的边界轮廓线, 并确定边界轮廓线 上每一点的坐标, 其中沿垂直边界方向的坐标为高度坐标。
17、 如权利要求 14所述的用于薄片型柔性体表面三维重建的方法, 其特征在于, 所述轮廓提取步骤还包括如下二个步骤:
•边缘检测: 利用边缘检测的算法进行边缘检测, 该边缘检测算法包 括 Marr边缘检测方法, Sobel算子, Robert算子, 拉普拉斯算子; ,高度提取: 扫描图像, 根据检测出来的轮廓边缘, 确定边界轮廓线 上每一点的坐标, 其中沿垂直边界方向的坐标为高度坐标。
18、 如权利要求 14所述的用于薄片型柔性体表面三维重建的方法, 其特征在于, 所述轮廓提取通过诺顿变换步骤来实现: 将图.像灰度值沿垂 直方向累积求和, 提取图像在垂直方向上的灰度分布; 对于水平方向的每 一位置, 由于沿垂直方向的该灰度的累积跟该水平位置处试样的厚度成线 性关系, 所以将该累积值除以一定的比例系数, 即可得到试样每一水平位 置处的表面轮廓高度。
19、根据权利要求 14所述的用于薄片型柔性体表面三维重建的方法, 其特征在于: 该方法适用于离线检测和在线检测两种模式。
PCT/CN2004/000681 2004-06-23 2004-06-23 Systeme et procede pour la reconstruction 3d de surface de corps lamellaire et flexible Ceased WO2006000123A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2004/000681 WO2006000123A1 (fr) 2004-06-23 2004-06-23 Systeme et procede pour la reconstruction 3d de surface de corps lamellaire et flexible

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2004/000681 WO2006000123A1 (fr) 2004-06-23 2004-06-23 Systeme et procede pour la reconstruction 3d de surface de corps lamellaire et flexible

Publications (1)

Publication Number Publication Date
WO2006000123A1 true WO2006000123A1 (fr) 2006-01-05

Family

ID=35781553

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2004/000681 Ceased WO2006000123A1 (fr) 2004-06-23 2004-06-23 Systeme et procede pour la reconstruction 3d de surface de corps lamellaire et flexible

Country Status (1)

Country Link
WO (1) WO2006000123A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8929587B2 (en) 2006-05-31 2015-01-06 Thomson Licensing Multi-tracking of video objects
CN109341527A (zh) * 2018-10-22 2019-02-15 广东工业大学 一种自动阴影补偿的结构光投影三维测量系统及方法
CN109993696A (zh) * 2019-03-15 2019-07-09 广州愿托科技有限公司 基于多视点图像的结构物表观全景图矫正拼接方法

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE19722607A1 (de) * 1997-05-30 1998-12-03 Michael F Braun Selbsttätiges Verfahren sowie Einrichtung zum Bestimmen der Eigenschaften einer Probe
CN1211319A (zh) * 1996-10-15 1999-03-17 Cis图像处理有限责任公司 纱线和织物模拟系统
US6542249B1 (en) * 1999-07-20 2003-04-01 The University Of Western Ontario Three-dimensional measurement method and apparatus
CN1534294A (zh) * 2002-06-06 2004-10-06 香港理工大学 织物表面分析方法及其系统

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1211319A (zh) * 1996-10-15 1999-03-17 Cis图像处理有限责任公司 纱线和织物模拟系统
DE19722607A1 (de) * 1997-05-30 1998-12-03 Michael F Braun Selbsttätiges Verfahren sowie Einrichtung zum Bestimmen der Eigenschaften einer Probe
US6542249B1 (en) * 1999-07-20 2003-04-01 The University Of Western Ontario Three-dimensional measurement method and apparatus
CN1534294A (zh) * 2002-06-06 2004-10-06 香港理工大学 织物表面分析方法及其系统

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8929587B2 (en) 2006-05-31 2015-01-06 Thomson Licensing Multi-tracking of video objects
CN109341527A (zh) * 2018-10-22 2019-02-15 广东工业大学 一种自动阴影补偿的结构光投影三维测量系统及方法
CN109993696A (zh) * 2019-03-15 2019-07-09 广州愿托科技有限公司 基于多视点图像的结构物表观全景图矫正拼接方法

Similar Documents

Publication Publication Date Title
CN1247994C (zh) 织物表面分析方法及其系统
JP7696692B2 (ja) 物体の連続的な画像をつなぎ合わせるためのシステムおよび方法
JP5337050B2 (ja) 車両用成形ガラスのひずみを反射された光学像により自動的に定量分析する方法
CN112798617A (zh) 类镜面物体的缺陷检测装置及其检测方法
JP4685971B2 (ja) 検査システム及び検査方法
EP3468163B1 (en) Intelligent internet high-definition scanner with laser correction
CN103196917A (zh) 基于ccd线阵相机的在线卷板材表面瑕疵检测系统及其检测方法
TW201033602A (en) Defect inspection device for formed sheet
JP2007532929A (ja) 分配システムにおける測定装置および方法
CA2508846C (en) Imaging device
CN113251926B (zh) 一种不规则物体的尺寸测量方法及测量装置
JP2013140050A (ja) 欠陥検査装置および欠陥検査方法
CN111602177A (zh) 用于生成物体的3d重建的方法和设备
CN1312461C (zh) 薄片型柔性体表面三维重建的系统和方法
CN113124771B (zh) 具有校准目标对象的成像系统
CN101932924A (zh) 信息处理设备和方法
JP5006551B2 (ja) 欠陥検査装置及び欠陥検査方法
BR112018071487B1 (pt) Sistema de captura de imagem e método para determinação da posição de uma estrutura gravada em um elemento laminado
WO2006000123A1 (fr) Systeme et procede pour la reconstruction 3d de surface de corps lamellaire et flexible
US20050033470A1 (en) System and method for the three-dimensional analysis and reconstruction of the surface of a thin flexible material
US20180189968A1 (en) Shape reconstruction of specular and/or diffuse objects using multiple layers of movable sheets
CN112683854B (zh) 物体质感测量装置
CN101501734A (zh) 用于凸起材料检测的方法与装置
CN208026651U (zh) 基于多通道光源的材料板形缺陷检测系统
US7570373B2 (en) System and method to measure parameters distribution in sheet-like objects

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): AE AG AL AM AT AU AZ BA BB BG BR BW BY BZ CA CH CN CO CR CU CZ DE DK DM DZ EC EE EG ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NA NI NO NZ OM PG PH PL PT RO RU SC SD SE SG SK SL SY TJ TM TN TR TT TZ UA UG US UZ VC VN YU ZA ZM ZW

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): BW GH GM KE LS MW MZ NA SD SL SZ TZ UG ZM ZW AM AZ BY KG KZ MD RU TJ TM AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HU IE IT LU MC NL PL PT RO SE SI SK TR BF BJ CF CG CI CM GA GN GQ GW ML MR NE SN TD TG

121 Ep: the epo has been informed by wipo that ep was designated in this application
DPEN Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed from 20040101)
NENP Non-entry into the national phase

Ref country code: DE

WWW Wipo information: withdrawn in national office

Country of ref document: DE

122 Ep: pct application non-entry in european phase