MXPA99005300A - Confocal system with scheimpflug condition - Google Patents

Abstract

Se describe un dispositivo para examinar objetos en tres dimensiones basado en un sistema confocal, en donde se proyecta la imagen de una rejilla (15) iluminada en la vecindad del objeto a examinar (10), la rejilla se encuentra inclinada con respecto al ejeóptico del sistema (12) por lo que la imagen proyectada se encuentra en el plano definido por la condición de Scheimpflug. Al enfocar este plano imagen sobre la superficie sensible de un CCD (16) utilizando para ello un divisor de haz (14) y garantizando la condición de Scheimpflug, se establece la condición confocal. Cada línea se encuentra enfocada a una profundidad en el objeto (11 a, 11 b, 11 c) por lo que al realizar un barrido en una dirección (20) y utilizando técnicas confocales de integración de la imagen de cada línea, se consigue integrar un mapa de profundidad del objeto en un solo barrido. Por medio de una calibración inicial se simula digitalmente una malla sobre la superficie del CCD con el fin de simular los diafragmas confocales e incrementar la resolución del dispositivo. El tamaño o frecuencia de la rejilla se ajusta con el f número del sistema para acoplar la profundidad de foco entre cada línea iluminada contigua.

Description

CONFOCAL SYSTEM WITH SCHEEVIPFLUG CONDITION BACKGROUND OF THE INVENTION The present invention is related to instruments that examine objects in three dimensions and that capture both the surface image and its depth map.

Among the techniques used to discriminate the depth of objects are those of the passive type, where it is not necessary to control the light source and those of the active type where the beam involved is usually laser light, it is controlled. Among the passive methods we find the techniques of stereoscopic disparity, camera movement, surface reflectance, texture gradient, occlusions, shadows and depth of focus discrimination among others [IEEE 1987, IEEE 1983]. Although these techniques are potentially powerful with difficulties in hardware and software that make them impractical for general use.

The most commercially accepted techniques are among the active ones, where the problem of the discrimination of the depth of the objects is solved by controlling the light that is made to influence the object. Among the most known methods of this type of techniques are interferometric ones that generally use laser light to measure small depth variations, such as those that arise due to deformations in the materials by tension. However, even when they are very precise techniques, they can not be widely used for general purposes because of their limited depth of field. Techniques based on the radar principle (measuring the travel time of a radiation) are used to measure large distances and provide moderate depth resolution information. For short distances triangulation methods are preferred and offer great potential to acquire depth accurately.

In the area of microscopy, we mainly find confocal techniques [Hamilton, 1982] that achieve a high lateral resolution and depth, even within semi-transparent objects, which is why they are widely accepted and find a large number of applications mainly in the area biological DESCRIPTION OF THE INVENTION The device of the invention presented in this patent is intended for general use, which makes the device susceptible to a large possibility of applications.

The invention is based on a confocal system which has been modified in order to simplify both the hardware and the software required in the analysis of three-dimensional objects. The features of the invention are clearly described below.

It is known that the lateral resolution of a system can be increased at the expense of the field of vision [Lukosz, W., 1966]. In a conventional confocal system this is achieved by projecting an image of point sources onto the object. In the case of this invention, what is projected are thin lines of light in order to increase the lateral resolution. To increase the field of vision, a horizontal sweep is performed perpendicular to these lines. In a conventional confocal system the light sources that are focused on the object are focused on small openings located immediately above some sensors, this is to ensure that the information they receive from each illuminated object point comes from these points and not from any other part. Consequently, in a system of this type the rest of the image appears dark.

In our case, when projecting a series of linear sources it is required that these are focused in turn on a rectangular CCD and that digitally a mask was generated on the sensitive surface of the sensor to perform the function of the confocal diaphragms. The digital mask must be adjusted to the image of the grid on the CCD so that the pixels of the dark areas do not provide information when integrating the image.

Figure 1 shows the principle of capturing the three-dimensional information of the object and the parts in which the invention consists. A condenser system (17) is shown at the top of the figure, 18, 19) that illuminates a high contrast grid with lines parallel to each other (15). As this grid lies in a plane that is not perpendicular to the optical axis of the system (12), the image of the grid that is projected by a lens (13), in the vicinity of the object (10), and which is also found in an inclined plane (11 a, 11 b, 11 c), this condition is known as Scheimpflug [WJ Smith, 1990]. This sets the position and inclination of the grid according to the region of the object to be examined. With this, each contiguous line of projected light is focused at a different depth. The image plane of the grid in the vicinity of the object is focused by means of a lens (13) and a beam splitter (14) on a rectangular CCD (charge-couple device) sensor (16) which is inclined in shape Similar to the grid so that each image line is focused in a different area of the sensitive plane of the CCD. In this way, only the light reflected from points objects that coincide with the illuminated lines is detected by the digital mesh of the sensor. The image of other points outside that plane are out of focus in the CCD, so they do not manage to significantly contaminate the confocally focused points.

As each line is focused at a certain depth in the object, slightly different from the contiguous lines, when making a sweep (20) the image that integrates each line only belongs to a plane of depth, thus integrating the information of all the lines of the grid we obtain the information of the topography of the complete object as well as its image. The thickness of the fringes and the f-number of the optical system determines the depth of focus that discriminate each line, which in turn allows coupling, adjusting these two parameters, the depth limit of a line with the upper or lower immediate limit of the contiguous lines. To obtain a greater efficiency in depth discrimination, the frequency distribution and the thickness in each grid area can be varied in order to compensate for the non-linearity of the longitudinal amplification.

The inclination of the grid and the size of the CCD determine the maximum depth Zo (figure 1) that the device can capture, the length of the grid lines determines the maximum width and the length (X) of the sweep gives us the length of the object to inspect.

Another feature of commercial CCDs that favor the operation of the device presented in this invention, is that either by its construction or mode of operation these sensors work as a mesh of very small detectors that coupled with the electronic mesh favor confocal detection of the object points, which increases both the lateral resolution and the depth discrimination of each point.

The integration of the image to obtain the information of the depth map of the object as well as its image is done using confocal image processing methods, such as the extended focus or the image in auto-focus depending on the application of the device [ T. Wilson, 1990].

An alternative of lighting is to use a grid reticular instead of a grid of lines, which combined with a similar digital grid in the CCD increases the lateral resolution. Another possibility is to replace the grid with a liquid crystal device to form the dark and transparent areas that make the grid function, with the convenience that its orientation can be adjusted in order to align it with the particular characteristics of the CCD. It also allows the limit to be taken in the dimensions of the illuminated and dark areas that are projected and that can reach to solve the CCD.

The fact that there is only one sweep in one direction makes the software and hardware involved easier, giving it better commercial characteristics than other confocal and other alternatives.

The trajectory of the sweep can be adjusted to a particular application, in order to obtain the most information of the object to be examined in a single step, this type of sweep goes from one in a linear and horizontal way, to any geometric shape that suits for reasons practices. The rotation or rotation of the system with respect to a point is applicable to robotic vision systems, stereoscopic systems, for sensors in industrial equipment, among others. Turns up to 360 degrees around a point near the grid or nodal point of the lens is convenient for a robotic vision system.

An initial calibration of the instrument allows that the alignment tolerances of the optical elements are not critical, allowing to mount these with a moderate precision.

STRUCTURE AND OPERATION In Figure 1 the lighting source of the device can be implemented using for example a halogen lamp (12) that coupled to a condenser lens (18) and a possible filter (17) (for cases where a certain portion of the spectrum is desired) of illumination) illuminates the linear openings of the grid. The capacitor is designed in a conventional manner according to the characteristics of the projection lens to ensure correct coupling. The grid can be made on a glass substrate by a conventional chrome tank such as those presented by the Rochi grids. The dimensions and characteristics of the grid are fixed according to the dimensions and characteristics of the CCD (16)For example, if a CCD XC-77 manufactured by Sony is used, a grid with a maximum frequency of up to 77 lines per millimeter should be chosen, so that the longest part of the CCD is the one with the inclination. The dimensions of the grid are similar to the dimensions of the sensitive part of the CCD.

The objective (13) that projects the slits of light while being in charge of focusing them on the CCD, the latter with the help of a beam splitter. This objective is chosen depending on the dimensions of the object to be examined, the object distance and the characteristics of the CCD. In any way, the system is required to solve at least the grid that is projected. For example, an effective focal length lens of 50mm, f / 5 focused on an object of 100mm in width and 300mm in depth.

The signal received by the CCD in a sweep is transmitted through a connection to a computer (21) which does the processing of the calibration and integration data of the images and shows the results of the sweep, for example, in the form of graphs and images on the screen, the computer can also control the displacement of the sweep of an electromechanical system (20) connected to it. This control can be fixed or it can depend on the results of the evaluation of a previous sweep.

An initial calibration of the instrument allows to determine the digital mask and allows associating a height and relative position of each pixel. This can be done, for example, by placing within the inspection range of the instrument an object that fills the entire capture range of the device and whose dimensions are known, such as an inclined plane, in such a way that it coincides with the plane of the image of the grid, in addition you can add some reference marks as a scale to improve the calibration. The digital mask is achieved when the image of the plane with the projection of the stripes coincides with the plane of the CCD, so this image is saved as a reference pattern, to be used in the processing of the confocal image of the object.

The technique can be adjusted to an infinity of types of optical systems for particular purposes, giving it a great possibility of applications among which we find the capture of digital images in three dimensions, robotic vision, metrology of objects in three dimensions, analysis of substances and internal structures of semi-transparent objects using the phenomenon of fluorescence, physical parameter sensors, etc.

REFERENCES Hamilton, D.K. etal (1982)., "Three Dimensional Surface Meansurement Using Confocal Scanning Microscope" Appl. Phys. B. vol. 27,, pp. 211-213.

IEEE (1987) Trans. Pattern Anal. Mach. Intel. 9 (4), 523-531.

IEEE (1983) Trans. Pattern Anal. Mach. Intel. 5 (2), 122-139.

Lukosz, W., (1966). Optical systems whit resolving power exceeding the classical limit. J. Opt. Soc. Am., 56, 1463-1472.

Minsking (12/1961), U.S.A. Number 3,013,467.

T. Wilson (1990), Confocal Microscopy, Academic Press.

W.J. Smith (1990), Modern Optical Engineering, McGraw-Hill, Inc.

Claims (7)

REGNVGNDICATIONS We consider as a novelty and therefore claim as our exclusive property, what is contained in the following clauses:

1. The principle of capturing the three-dimensional image and the way in which the device that examines objects in three dimensions works, understanding as a device the configuration of the elements and their parts as a result of the application of the principle.

2. The capture procedure to obtain the information of the three-dimensional object and the handling of the data.

3. The application of the condition of Scheimpflug together with the confocal principle as part of the principle of capture and discrimination of the depth of the image determines the orientation, inclination and position of the planes objects and images of the system and of the other elements of the device.

4. The procedure of calibration and adjustment of the device

5. The lighting system that generates areas of light and shadow (grid) on the object to be examined and that is on an inclined plane and that increases the resolution of the image that is captured when scanning, which facilitates the software and hardware involved.

6. The characteristics of the grids, with thin lines of high contrast either with variable or constant frequency.

7. The option to use a grid with a reticular shape. The option to use a liquid crystal element to form the grid. The use of a detector or sensor of the type (CCD) or similar technology of rectangular shape The way in which the CCD signal is handled to simulate a confocal diaphragm sensor mask. The sweep path options of the system and that due to a particular application fits a particular geometry. Such as linear sweep trajectories (in any direction), turns with respect to a point, undulations, among others. The use of a type of optical system in particular image-forming as a confocal objective for a particular application. Such as image relays, telephoto, telescope objectives, microscope objectives, collimators, etc. The following applications of the device: a) As an instrument for measuring the topography of objects. b) As information capture devices that contain three-dimensional objects, whether transparent, semi-transparent or opaque, such as dimensions, textures, colors, fluorescence, brightness, structures or any other physical parameter relative to the object to be examined. c) As robotic vision devices. CONFOCAL SYSTEM WITH SCHEIMPFLUG CONDITION SUMMARY A device for examining objects in three dimensions based on a confocal system is described, wherein the image of a grid (15) illuminated in the vicinity of the object to be examined (10) is projected, the grid is inclined with respect to the optical axis of the system (12) so the projected image is in the plane defined by the condition of Scheimpflug. By focusing this image plane on the sensitive surface of a CCD (16) using a beam splitter (14) and guaranteeing the Scheimpflug condition, the confocal condition is established. Each line is focused at a depth in the object (11 a, 11 b, 11 c) so when sweeping in one direction (20) and using confocal techniques to integrate the image of each line, it is possible to integrate a depth map of the object in a single sweep. By means of an initial calibration, a mesh is simulated digitally on the surface of the CCD in order to simulate the confocal diaphragms and increase the resolution of the device. The size or frequency of the grid is adjusted with the f number of the system to match the depth of focus between each contiguous illuminated line.