JP2000500236A - Computer stereo vision system and method - Google Patents

Abstract

(57) SUMMARY The system and method of the present invention are intended to allow a robot, device, tool, etc., to view its operating environment with a pair of identical cameras. Aligned, compatible, coordinated with its field of view (field of view), parallel and adjusted for parallel deflection (M) between all lines in the field of view, and the system and method of the present invention. , Robots, devices, tools, etc. are intended to be able to immediately identify everything the camera sees at the shooting speed. With this system and method, computer vision receives images from a camera to a designated location for this purpose, and a backup for the field of view is created from the images stored in spatial memory. Calculate the distance and the shape of each point for each point in the image, and register the various features. The system consists of a memory register for the calculation of each motion, motion, speed, and direction of each shape in the image of the main camera, which has been received after being filtered and color-filtered at predetermined time intervals, the basic shape. And registers for data tables such as "intrinsic" tables. This register is intended to detect data, features, definitions and draw conclusions for key elements. The data, features, definitions, and conclusions can constitute a key for the unidentified shape that is compatible with the register of the recognized saved shape, thus completely and substantially identifying the unidentified shape. Perform immediately.

Description

DETAILED DESCRIPTION OF THE INVENTION Computer Stereo Vision System and Method Field of the invention The present invention relates generally to computer vision and identification viewed in real time by means of cameras, computers, etc., and 3-D computer vision with two or more "stereoscopic" cameras. Background of the Invention The demand for computer vision is to solve problems in all areas, such as research, astronomy, robotics, industry, agriculture, manufacturing, services, driving, security, assisting the blind, detecting phenomena, etc., and for the actual use by humans. Is a daily necessity. Today, as both internal and external computer memory grows, the physical dimensions of computer components are gradually decreasing and there are various types of high speed processors. Therefore, if we could create a system and method for computer vision that could identify everything we see in real time, everything would be easier and simpler. current situation Computer vision using one or two cameras that exists today focuses on individual, defined, known and predominantly static object visions, their identification is redundant and comparative Only partial, focused and focused, does not analyze and identify everything seen at the same time as the image is taken, and requires many devices such as sensors, lighting, measuring gauges, It is cumbersome, limited, inefficient and does not give a satisfactory solution. Purpose of the invention It is an object of the present invention to be able to analyze and identify all visible shapes simultaneously and at the speed of photography, by means of a camera connected to the computer, computing the dimensions, movements and other perceptual features. To provide a system and method. In addition, it is an object of the present invention that all automated devices (tools, computers, robots, etc.) can view and identify all tangible objects and phenomena by means of various appropriate visions. To make it possible. Enabling movements, tasks and tasks performed by approximately everyone to be more accurate, efficient, faster, better, anytime, anywhere, physically difficult to access, dangerous, inaccessible, boring, Help and execute. Further, it is an object of the present invention that a person can use the multimedia of the present invention to display a space around a specific area from a remote place in a 3-D display by using the multimedia software with the means of the present invention. , Seeing by using glasses and / or special equipment designed for 3-D vision, or lines and regular monitor screens for data transfer. It is an object of the present invention to make it possible to construct devices therefor using equipment, systems, circuits and electronic components, basic software, etc. that exist in some form on the market. And by their combination, combination, addition, extension, etc., the device can be created and / or assembled in accordance with the present invention, including its ability to adapt to the environment. The average electronic engineer, computer engineer, system analyzer, etc. can design, assemble and configure the device according to the present invention. The present invention The present invention comprises an innovative system and method for preferred stereoscopic computer vision, including and operating as follows. Introduction 1. Overall, all materials as software descriptions, which do not form part of the process of the present invention, are presented as samples only. Its purpose is to explain the process of operating computer vision, but not to diminish the essence of the invention. 2. Due to the existence of computer vision systems of various types, models and sizes (eg, large, small, medium, Philips screwdrivers, etc.), depending on the data that computer vision must transfer to serving customers, The description is general, but the basic operating principle in all of them is the same. 3. Computer vision is a means of providing "seeing" services. The challenge is to look at sections in space, decrypt as needed, store whatever is needed in registers, and transfer the viewed or related data. If necessary, transfer data in 3-D format or "multimedia" (including data on the shape, location, location, size, color, distance, etc. of the object seen), all for the purpose of the computer vision system It is also possible to design 4. Because computer vision is thought to serve customers and customers are likely to perform limited operations, their workspace is usually limited. Therefore, a computer vision system tailored to such a customer must provide the customer with the necessary data to satisfy the task. With the help of the appropriate data and software, it becomes possible to "get familiar" with the environment in which the customer operates and recognize the shapes of objects that can be found in the area and whose perception is important to such customers. In addition, the system recognizes by the degree to which customers must meet and handle in all disciplines, and the equipment and tools to use in their mission framework. 5. Computer vision can use additional auxiliary equipment, such as a memory system, to help customers make measurements. Also, such auxiliary devices can be external to the system or customer. There may be interrelationships, compatibility, references, consultations, interoperability, data transfer, etc., between the customer's computer vision and auxiliary equipment. List of drawings (figure) Drawing 1 is a schematic block diagram-like sketch of a preferred system with possible connections according to the invention, with two processors. Drawing 2 is a schematic sketch of the parts that make up the system constructed and operated in accordance with the present invention. Drawing 3 is a horizontal cross-sectional view schematically illustrating the plurality of central viewing axes, the field of view and the parallel constraint lines of the field of view and the visual image of the system shown in FIG. Drawing (Figure) A / 3 (× 4 magnification) is a parallel visual image of Drawing 3. Drawing (FIG.) 4 is a horizontal sectional view schematically showing the field of view (of the system described in FIG. 1) and the shape seen at various distances and various sizes. The drawing (FIG.) A / 4 (magnification: × 4) shows a view of the shape shown in FIG. 4 viewed simultaneously by both the left and right cameras of the system described in FIG. Shooting and camera 1. A pair of identical cameras 3 (FIGS. 1 and 2) are arranged so that the shooting angles including variable and enlargement / reduction are arranged on coordinate axes and coexist on a common plane (hereinafter referred to as a horizontal plane). A parallel visual field (FIG. 3) is formed. That is, the distance of each site line to the field is the same for both cameras from (0: 0) to (Y: X), and is a fixed distance M (FIG. 3 and A / 3) at all shooting distances. is there. The vertical movement of the camera (perpendicular to the horizontal plane) is constant and identical. Pictures taken by the camera are translated into a computer language (digital) by the camera or other equipment and then received in the input memories A / 51 and B / 51 (FIG. 1) according to the shooting speed and / or the like. You. Adjustments by the computer vision system can be made physically during installation and / or at any other time. 2. The two or more cameras in one enclosure or separate packaging are similar to a video camera, such as a CCD camera, and integrate to convert the received data in the image to digital data Including one or more of the following: a) Various speeds and any light, such as IR, visible, etc., by any means such as light amplification, and any such as insufficient B) magnification or reduction device including telephoto or microscopic means c) linear, concave or convex, optical image 25 with desired resolution (FIGS. 3 and A / 3). 3. The camera is equipped with the necessary auxiliary equipment that works according to the instructions given by the computer vision system. a) Adjust the lens (distance, light, etc.). b) Move the camera (lateral, circular, upward, downward, etc.). c) Set (telescope, microscope, starlight amplifier, etc.) d) Serve the computer vision system and camera and bring them to a situation where computer vision is performed in a proper, desired and accurate manner at a specific time. To go. translator The A / 59 and B / 59 translators (FIG. 1) can form an integral part of the camera and / or from operation, measurement, coupling to external equipment and / or any attached means It can be translated into a computer vision system and vice versa. It can serve all or only part of the system, or some of the above, to a computer vision system. Image input process 1. Any image (including a color image) received from any of the cameras enters the input memory A / 51 and B / 51 (FIG. 1) as it is at the designated position, creating a space proportional to the camera resolution. Occupy. The images of all the cameras enter all separate locations of the input memory, such as an image simulating memory device (eg a “screen card”), at all shooting speeds with all the colors. After the color filtering 53 (FIG. 1), it enters the motion identification register 54 (FIG. 1) at predetermined time intervals. 2. The transfer of the image to its location in the spatial memory A / 52 (FIG. 1) to one camera or two cameras in the case of 3-D photography, if the computer vision system is part of the user Register all images received from a fixed, scaled and / or other image according to the Consumer Movement 6 (FIG. 2) and / or according to the camera movement on the horizontal and vertical axes I do. The image is saved as an outline image in the horizontal and vertical coordinate axes and is constantly updated. The horizontal and vertical coordinates of all images from each and all cameras are updated according to the motion, motion calculations. They are also updated according to the movement of the space counter 57 (FIG. 1). 3. The image received from the camera according to the shooting speed is scanned by the A / 55 processor (FIG. 1) and during the scanning process, for each unidentified or moved / advanced shape in the image, the image from the camera is taken. The distance and other data that can be calculated and / or resulting from the image, such as color, are calculated, and the image is transferred to the appropriate location in the spatial memory register and to another location, such as a consumer. Scanning and matching methods 1. Assuming that the left-to-right and top-to-down scanning starts from the right camera (main camera) image to pixel (0: 0) and then to the left camera image pixel (0: 0), Until a point (Y: X) in all camera images is reached, scanning of image data according to shooting and / or other speeds is performed. If a 3-D display is required, the scanning data of both each and all cameras, not just the main camera, are separately matched to the spatial memory image according to the coordinate axes and the spatial counter, respectively. Is updated and a match is found between the resulting image and the image. 2. The modified image and the dots therein must be matched for distance measurement and / or contour identification purposes. In addition to normal matching, they are matched with each other. 3. Scanning is performed for the purpose of matching and calculating the distance L (FIGS. 4 and A / 4) to point D in the horizontal line X and column Y in the images of both cameras. X1 is the number of pixels from a point (Y: 0) in the main camera, and X2 is the number of pixels from the point (Y: 0) in another camera to the same dot. The following differences are obtained: Wr = X2-X1. This allows the calculation of the distance. 4. Pixels indicate points seen by the right and left cameras, which are located beyond the matching distance (2M and pixels indicating O = Wr <＜). A change in distance cannot be detected beyond this distance. The same point in the images received by both cameras is located at the same position (Y: X) and the pixels representing them are the same. 5. From the matching distance, the shorter the distance, the greater the distance Wr between the pixels. That is, moving from a distant shape to a closer shape increases the distance Wr between pixels, while moving from a closer shape to a distant shape decreases the distance Wr between pixels. as a result There is a dot in the image of the other camera that does not appear in the image of the main camera when it gets farther and closes up, but the computer vision system takes it into account when performing the matching. These depth dots, if continuous, also form contours and / or dividing lines between different regions. 6. At first look, the vision computer blocks additional images, except for the spaced motion detection registers, and performs pixel transfer scanning and matching of the first seen image. Computer vision systems match pixels (0: 0) in the image of the main camera with the same pixels in other cameras. If there is a match, the difference is 0 = Wr, and the system moves on to the next pixel in both images to be transfer matched and continues until a non-match is found. If there is no match, the difference is 0 ≠ Wr, and the computer vision system first matches the pixels of the image of the main camera one after the other until it finds a match with the same pixel of the unmatched image of the second camera. Continue the transfer and matching, or until the difference is 0 = Wr, and if no match is found, the computer returns to the same unmatched pixel of the image of the primary camera and the image of the second camera. Pixels are successively matched until they are matched using the definition of region limitation based on spaced motion detection. 7. Successive representative pixels on the same line in both images at the same distance are sequentially matched in order, and if two pixels do not match, it is assumed that one of the points has changed in distance, and accordingly The matching process for the calculation works. 8. In a standard matching screen, if there is no match between the input image and the spatial memory image, the computer will start matching between images received from the same place where there was no match in each image, and the previous information , Detect whether any shape has moved, changed, or is new, and calculate the distance of each point in all matchings. This is done until a match between the received image and the spatial memory image is found and proceeds as normal. Distance calculation process 1. For the purpose of calculation and image handling, the coordinates 57 can be assigned to a counter, for example from (0,0) to (Y, X) or (−Y, −X) to (Y, X) (FIG. 1) (Showing a line), which is a kind of grid on which the spatial memory image is placed. For example, the "0" point is the center reading point Op (horizontal point in the image) on the horizontal level for each camera (Figure 3 and A / 3) and the camera is leveled and "North" of the compass. Turned north according to the direction. If the camera's eastward lateral or upward movement is positive and the westward or downward movement is negative, the computer automatically (eg, angle of one pixel) according to the movement of the image by an external element or compass or auxiliary system. Updated (by size) and accept camera data accordingly. One coordinate indicates a peripheral horizontal point and the other a peripheral vertical point. 2. For the purpose of region orientation, for example, there is a three-dimensional counter 57 (FIG. 1) associated with the center line of the field of view of the main camera (center point of the camera). For example, a counter (1) that counts the progress of the camera movement from “East to West” from the set reference “0” point, “North-South” from the set reference “0” point A counter (2) for counting the progress of the camera movement, such as a counter (3) for counting rise and fall (height) for sea level or any other established level forming a "0" point. is there. 3. For all images received into the input memory, a calculation is performed to match the counter to coordinates with respect to the first identified point therein. The resulting image is transferred to a spatial memory image according to calculations and relative coordinates. 4. If a video camera (CCD) or any similar camera is used, the rotation of the camera on a horizontal or vertical plane maps the peripheral image (in standard shooting) with coordinates according to the number of pixels in the camera image. It is possible to do. Thus, the space is mapped by coordinates that indicate corners within the envelope of the sphere. The center is the angle point of the shooting angle of the camera. When one revolution is completed or when the end of the shooting / visual space along the horizontal or vertical axis is reached, the computer vision identifies, for example, the number of pixels / coordinates and the course or motion / movement in the motion "Knows" that you have to return to the beginning coordinates of the space. 5. For example, during zooming, the camera zoom is activated to reduce the angle of the captured image, and the coordinates added accordingly are decimal places, for example three decimal places. 6. The number of pixels occupied by all shapes in space depends on the number of pixels in the optical image, the physical dimensions of the shape, its proximity to the camera and the open viewing angle. The larger the number of pixels, the smaller the viewing angle aperture of the optical image, the more pixels will exhibit the same size shape (at a fixed distance from the camera) and the better the shape identification. Furthermore, the larger the enlargement, the larger the number of pixels having the same size shape. Method of calculation The calculation of the distance L from the camera to the matching point to be calculated is based on the distance Wr, the identification, the physical dimensions and resolution of the optical image, the matching between the cameras and the constant distance M (parallel separation). 1. Radius r from the angle part α of the viewing angle to the optical image ₀ , Width X (on X axis) ₀ , And the formula for scaling A is: r ₀ = 360 * A * X ₀ / 2 * π * α (the angle around the full circle is 360); and K ₀ Are all constant data in the formula: K ₀ = 360 * X ₀ / 2 * π. 2. The radius r to any point seen from the viewing angle α, the number X of horizontal pixels (on the X axis) in the optical image and the formula for calculation are: r = 360 * A * X * M / 2 * Π * Wr * α and K are all constant data in the formula: K = 360 * X / 2 * π. 3. K ₁ Is a constant in each case and is directly proportional to the enlargement / reduction A and inversely proportional to the viewing angle α, and the equation is ₁ = A / α. 4. The formula for calculating the distance L is: L = K ₁ (K / Wr-K ₀ 4.) (Factor of variable Wr) Indirect distance L ₁ Equation for the calculation of, where Q ₁ Is the size of what the pixel indicates at a location of a shape having a standard size, such as the height of a known person. The calculation formula is: L ₁ = 360 * A * Q ₁ * X / 2 * π * α-r ₀ = K ₁ (K * Q ₁ -K ₀ Therefore, the result: the calculation of the distance to the point makes it possible to calculate the size of what the pixel indicates for the point it indicates. The calculation is based on the frame of the area to which the point belongs, and other information, such as the width, height, depth, size, angle, features, and states (fluttering, fluttering, etc.) that require the calculation. Assist / allow detection of definition data. Process of area division and data collection Each of the regions 11 and 12 (Drawing A / 4) is given a code (temporary name) for identification. The region can form other regions of various dimensions therein. Each area has a normal contour (except for dots and lines), a fixed distance of the area, a difference from the environment or a different color or contour and / or a separation between the area and the surrounding area, a movement of the area and It must have characteristics such as the state of the area (trembling, hanging, etc.). These, as well as other data, are identified and defined by a computer vision system with a B / 55 processor (FIG. 1). 1. Color Separation 58 (FIG. 1) Software 56 (FIG. 1) processes the colors of the received image, separates the colors, uses other prior region definitions to help define the region, and governs the region. For each of the target colors, an additional definition of that color among all the colors in the region is detected. For example, four types of divisions are possible. a. The amount of color: 100% -90% of the color in the area; b. The amount of color: 90% -70% of the color in the area; c. The amount of color: 70% -50% of the color in the area; d. Color amount: 50% or less of the color in the area. In addition, I also want to emphasize aspects of color diffusion (diffused dots, spots of a certain size, etc.). 2. Simultaneously with the reception and in addition to the normal reception, the image of the main camera is passed through the color filtering 53 (FIG. 1) (save source) and saved in the motion detection register 54 at predetermined time intervals (FIG. 1) and fixed. In the scaled and desired and known scaled photos (they can be several), they are duplicated and examined at equal time intervals. Based on prior data (eg, colors, shadows, calculations, etc.) taken from the received image, the motion, motion, speed, angle (angle for the viewed object, camera of the viewed object and others in space) Angle), as well as motion within the region (such as eye movement), and movement within the envelope of the region (position, hand movement, etc.) for the regions / shapes 11 and 12 (FIG. 4) and / or calculation. Whether the base is stable and the top is moving (eg trees), whether the movement of the area is flowing (eg rivers), the direction (north, south, east, west, up, down, forward, backward) , Right and left), speed, and the state of motion and / or motion type and shape, and the position and area and shape of the computer vision that could be checked and not detected based on the spatial counter It can help to define the framework. 3. Registers of the basic and essential basic shape C / 52 (FIG. 1) (eg geometric ones) have contours with lines and / or a few dots around them-in black and white-computer vision, the challenge And in the simplest form compatible with the purpose. This basic shape is based on the input shape through appropriate frame processing to obtain (schematic) comparison data on the processed shape (one shape can match several basic shapes). Saved in an order that allows immediate access for comparison in size matching on shapes. For the definition setting of the area contour, the memory for the basic shape (in principle 256 or less), such as rectangle, grid, dome, point, etc., depends on the number of saved shapes B / 52 (FIG. 1) (number 100,000). 4. Tables 58 (Drawing 1), such as the "intrinsic" tables, are compatible with the computer vision and may be included next to / into one table for area / shape data contained therein and / or obtained from one table. There are other tables, some of which are this table and some of which are the following tables, which use the data in a particular order for feature detection, definition, and conclusions. All or any collected features, definitions and conclusions are added and / or participate in key elements for detection, matching and identification. Additional features, detection, setting of definitions and derivation of conclusions are possible, e.g. immobile objects, vehicles, floating objects, creatures, e.g .; whether an area is between one size and another, Here each table shows possible movements and / or movement treatments, possible color treatments, whether it is moving and whether there is movement and / or movement in the envelope surface and / or therein. Covers subjects such as handling changes in distance from the surrounding environment. 5. The data table 58 (FIG. 1) for the re-presented shape, found at the photographic distance adjusted to the stored shape, and its presence in the image are very appropriate. In these cases, immediately after receiving the image and calculating additional individual data, such as size and color, the computer vision system finds that the shape has reappeared, and it detects and stores the stored shape against the registers. Construct a key element for identification. 6. If required, necessary and possible, identification data 57 (FIG. 1), such as heat, radiation, voice, taste, smell, etc., are also added to the existing identification data. 7. Internal factors (spatial counter, compass, etc.), external factors (eg, from camera, zoom, etc., from user, eg, speed of movement, etc.), printers, speedometers, and / or accessories of any system, eg, compass From telescopes, telescopes, rangefinders (hereafter ancillary means), and other relevant data, supplementary data 57 (FIG. 1) can be obtained in accordance with computer vision requirements, desired speed and computer language (digital). 8. A stored and recognized shape B / 52 of image, map, sign and / or text data near / on any shape that can be defined and / or given a name and / or independence identified. (Drawing 1) (eg, immovable objects, plants, creatures, backgrounds, signs, phenomena, and other things) are stored in any of the following four shapes in a specific and known order that matches the key elements: Is done. a. B. including color, as received by camera B. Black and white image S. Simplified shape (sketch) In the form of a table together with names and data, or in a stated shape, such as a card index. Images and maps are in photographic standards, which depend on shape size and shooting distance. The shape, stored in the form of an image, can be stored in several places-at each place it is saved from a different side view angle, so that it is identified from all directions. For example, for a 3-D image, six different images can be stored-each in the Cartesian direction. Process of recognizing area in image and matching with stored shape 1. Since the stored shape cannot represent all of the possible sizes of the shape at various distances in space, a "logarithmic" system and / or other forms of processing 56 and B / 52 ( FIG. 1) performs size matching between the shooting distance at a predetermined moment in the space of the actual shape received in the image and the appropriate shooting distance of the stored shape. 2. One region may include additional regions within it. Matching and identification starts with the largest region and progresses to smaller regions. If the interior region is part of the data of the identified shape within the region, more generally, identifying the region eliminates the need to identify the interior region, but completes the identification. If necessary (eg, position), identification is performed. 3. The key elements 58 (FIG. 1) from the data, features, definitions, and conclusions arranged in a particular order adjusted to the same order as the database of known stored shapes quickly classify the collected key elements. And arrangement, whereby detection, matching, and identification of an unidentified region can be performed. In normal field of view, there is very little unidentified area at any shooting speed, so identification is almost instantaneous and detection is done like looking up a word in a dictionary, i.e. the category is language and the key element The order is alphabetical, with the whole word being the key. A component may be absolute, in-between, or possible. For registers of recognized and stored shapes that contain data about the shape, such as the form of images, maps, signs, and / or textual data (in the form of a table, near or above) stored in the register. If relevant to images and maps, depending on the physical dimensions of the shape, the shooting distance and the purpose of the computer vision, the identification by size and shooting angle according to the stored shooting standards for the shape Is performed. 4. Any shape that requires a deeper analysis of secondary areas, parts, secondary details, age, form, location, etc. will be processed according to its needs and any vision required to be provided to the serving customer by Vision Provides more specific and detailed data, including data. Signature detection and identification Software for detecting and identifying salient signatures in images. For example: The software program detects, matches, and identifies image signatures (e.g., marks, documents, road signs, traffic lights, etc.). According to the identification data, the software knows, for example, where in the shape the document or predetermined sign is (eg the location of the vehicle registration number), and matches, identifies, and sizes the stored document or sign. Know to access that place to do the matching. 2. To assist in the detection process, there is a catalog or dictionary [signs, documents (words or names in local language or English, etc.), formats, road signs, traffic lights, etc.]. Partial vision 1. If a part of the shape is hidden, a partial vision arises and the entire shape must be detected and identified based on what is visible. 2. In such a case, computer vision detects the part that appears in the image and displays for this part its particular shape and all details about its visible part. Unidentified shape If no matching shape is found in the image during the matching and identification steps, Save the shape along with all relevant data to the assigned location in the register (area, part, detail, etc.). 2. Computer vision systems count how many times this shape is encountered. If this shape is not encountered again within a reasonable time, the system uses that location in the register for other similar cases. 3. The computer vision system may include, for example, a facsimile interface, a monitor, and options for transmitting to or receiving from external components, such as printers 31, 32 (FIG. 2), voice systems, computer communications, and the like. Such options are used to identify shapes that could not be identified according to system memory. Map analysis A computer vision system that serves devices moving in buildings, urban areas, fields, suburbs, etc., can analyze the images or courses viewed and transfer the data to the same locations in the map stored in system registers, Can be compared exactly. 1. The computer vision system finds its position on the map by knowing its place of departure on the map or by identifying data from the actual image, either by a space counter or from external factors. For example, the system reads the name of the street, and the name of another street that intersects it, and the number of the two houses on the same side of the street, and identifies the location that includes the located urban area with an appropriate plan. . 2. Maps provide data on courses and courses. a. Data on obstacles on and off the course, pedestrians crossing, railways, intersections, intersections, slopes, traffic lights, road signs, etc. The data may appear on drawings, maps, etc., some being adjacent to the course, and others being shown in other ways. B. The data can be registered in the form of a document, a code indicating a shape, or a data table (for example, a document, a code, a sign, a traffic signal, and the like). Rotational motion The computer vision system may be configured to move the shape being viewed, and / or to tilt the camera up, down, and sideways, and / or by rotating the user about an axis, and / or It is assumed that the movement of the user connected to the camera changes, so that the user sees a changing space. The spatial memory image covers all angles of the circle and the desired field of view of the computer vision system. Each time the camera receives an image at an angle, the system detects a change in the matching between the latest image and the image stored at the appropriate coordinates in spatial memory, and determines the location and location of the latest image. The different part is replaced with a part of the image stored in the spatial memory. Storage and display of space in 3D and multimedia From each pair of computers, the computer vision system stores a complete spatial memory image. 1. The images are provided as a whole in a circular motion, and each image is stored separately, creating a complex spatial image around each camera. 2. The aerial image allows the computer vision system to translate the data of the space it has moved or found into any 3D display software program (including necessarily the details of the computer vision system, including identification details and display of the data) against any other factors. (Which does not form a part) can be transported, displayed and displayed. 3. A display made possible by connecting an external display or other type of display device, which allows the space around the computer vision system to be viewed in 3D as if it were real. 4. Data can be transmitted to the screen or other type of display by interface 33 (FIG. 2), either by physical transmission or transmission. It does not form part of a computer vision system, but does form part of a user or user system supplement. Area / shape tracking 1. At any time, the location and location of the region / shape, and / or movement and / or movement, and / or part of the shape in the envelope and / or within the interior (eg, moving objects, growth, etc.) To track the change, refer to the coordinates of the first point scanned for the region / shape in the image of the main camera and the coordinates of the impending point in the envelope and / or the coordinates of the shape indicated by the region. Thus, for tracking purposes, the unidentified region is saved under a temporary name, and the identified form is saved under that name. 2. All data gathered from video, calculations, motion identification performed at predetermined time intervals, and a database of stored forms, and all identification means associated with the shape, until the shape disappears from the camera's spatial memory image Stored for your needs. In the register Storage If necessary, the computer vision system sends the image (full image, sample image, partial image, etc.) or shape (image, graph, signature, code, etc.) to a register for later use for storage I do. The computer vision system must save or store the image so that it can be later restored when needed, in the following manner. 1. Complete and continuous images and / or data defining an image from one moment to another can be stored in appropriate registers (eg, to understand processes, evidence, etc., of the things that one wants to see, (Before, during and after recording). 2. Updating and replacing the image or map in the register with a newer, updated one. 3. Process and / or data and / or shape as seen by all methods. External connection 1. The computer vision system has connection means 50 (FIG. 1) including a computer, robot, user, device 20 (FIG. 1), printer, fax machine, controlled device, microphone, loudspeaker, A communication line or the like can be connected. Of course, all computer vision systems only need to have the necessary connection means (one input / output, one input / output, And there will be a first code for opening or tuning). 2. Coordinators A / 59 and B / 59 (FIG. 1) coordinate connections between the computer vision system and cameras, users, etc., and from any of them to the computer vision system. transfer After identifying the details and data according to the purpose and requirements of the user for whom the system is intended to serve, the system transfers the required required data to the user by a suitable transfer means or interface. 1. Based on the received images, calculations, information stored in memory, processing, decoding, additional data, and various processing, the computer vision system provides data to the user that the system serves, and All direct and computer processed data (eg, measurements, distances, locations) are presented to the user. Reports can be continuous or on demand. 2. Computer vision systems can identify images globally and in detail, and can identify the location of each and every detail in an image. Also, if desired, the requested image or the entire image can be transferred in the form of instructions, documents, speech, fax, etc. along with various forms of detailed analysis. 3. The user receiving the report from the system needs to be prepared to accept and process the received data to operate accordingly. When the user uses the computer vision to perform a predetermined operation, the user can use the system for accurate operation, instructions, and the like. 4. The spatial memory image is transferred directly to the user directly, or from the spatial memory as needed, to a monitor screen or suitable device for 3D display and / or with identification data as needed. Knot 1. The computer vision system shall be protected as much as possible from any kind of dazzling light, flash, etc., including lasers, etc., in any location and from other physical damage. 2. Computer vision systems are compatible and can operate according to user needs and requirements. 3. The computer vision system includes a software program 56 (FIG. 1), a software program that adjusts the size and shooting enlargement / reduction data according to the standard size and shooting angle of the stored electronic and general circles and stored shapes, and a 3D display. Software and multimedia, as well as other software programs required. 4. The received image, data, computation of the image in spatial memory, and other information about the received image or map or data stored in all memories of the system may be data, data only, and / or input or spatial stereo. The images, including the images, may be sent directly to the user such as a robot according to the request and design according to any method, and the user can extract the data when necessary according to the system design. 5. All of the above operations are performed by processors A / 55 and B / 55 (FIG. 1) and, if necessary, additional processors and processing and computing means. These use computer equipment, components, one or more electronic circles, etc., and combinations thereof, required and adapted for computer vision purposes. In these roles, each part, part, electronic circle, system, and the like plays a role without interfering with each other, and there is comprehensive coordination and compatibility among all of them. For example, the processor No. The processor No. 1 has completed its part of the operation A, and the processor no. The processor No. 2 informs the user of the fact that he / she may continue his / her own part in the complicated operation A by means of parts, an electronic circle, or the like. 2 is notified. Commercial implementation 1. Computer vision can be used as follows. a. As a viewer (monitoring device, analyzer, decoder, reporter, etc.) b. 1. As a viewer that collects data by any means and saves it in a database or register (complete, partial, classified by any form, etc.) Due to the many possibilities of computer vision, several standard computer vision systems can be used, each of which is adapted to provide certain services and perform certain tasks, Therefore, it can be used for “protection”. On the other hand, special jobs can therefore be used for “protection”. On the other hand, for special requirements, we design computer vision systems tailored to special applications. 3. Each computer vision system is provided with standard and special abilities. 4. The computer vision system can be tailored to the special uses and needs of the designer or a user who is a constituent in the design phase, and the system can be incorporated as part of the user or as a separate unit serving the user.

[Procedure for Amendment] Article 184-4, Paragraph 4 of the Patent Act [Submission date] July 2, 1997 (1997.7.2) [Correction contents]                                The scope of the claims 1. A stereoscopic computer vision system, comprising:   a. A pair of identical, aligned, and co-located cameras at a fixed distance M The camera has a shooting angle and an enlargement / reduction arranged on a coordinate axis. For a shadow camera, at a fixed distance M at any shooting distance, from (0: 0) Match the field of view and coordinates of the same site line in both cameras until (Y: X) And form an aligned, optically parallel field of view, the image of the camera Received by the input memory device in the computer, translated into computer language,   b. Based on step A, the pixels of the received image may be captured at the shooting speed and / or At other speeds, by available means, according to coordinates and space counters, Matches each of the previous images in the system's spatial memory registry If there is no match, update the system and enter the two Advance image matching,   c. On the basis of steps A and B, an unidentified area, that is, a certain movement is performed, For an area that is viewed simultaneously by two cameras, the system will determine the type of image, such as color. The camera searches for various features and captures them from the start of the line to its matching point. Based on the difference in the number of pixels Then, the distance and size of the point in space represented by the pixel are calculated, and And calculate the area size, distance, etc.   d. Based on the steps B and C, the points of the unidentified area and the registered images are mapped. When changing the distance between the camera and the surrounding environment and the distance between the two cameras, Therefore, there are points around the area that do not match, and these depth dots are Create contour lines and / or parting lines in addition to distances approximately equal to For regions, changes in color or darkness between regions and / or changes in region motion / motion The frame or outline of each region is detected and defined by available means. And can   e. Calculation, matching, definition setting, conclusion, depending on the available means One camera to collect additional data that allows the derivation of The video received from the camera is filtered by a color filter and copied for a certain period of time. Enters the motion identification register of the remote and is inspected at regular time intervals. Then, the frame or outline of the detected area is set to the reference shape register and size. And the data from this register allows, for example, It is possible to calculate the size and the movement / movement speed within a given time, additional data Data, part of the data entered, and part of the data previously calculated / detected For example Matching against registers in tables such as "intrinsic" is possible, for example, Object, vibrating object, moving object, for example, The size has changed from one size to another, and within the envelope and / or For other key elements, such as whether movement and / or movement is taking place in the area, Set definitions, draw conclusions,   f. Derived from ordered data, features, definitions, and conclusions By the key element, the register of the registered shape is provided. The area can be detected, matched and identified by the means   g. Unidentified regions are stored under a temporary name, and identified regions are identified as envelopes and / or Location, position and / or motion and / or motion within the shape In order to always track and monitor such changes, the first Refer to the coordinate data when scanning the dots and additional dots, and 3D computer vision system stored under the name of the identification area itself. 2. The computer vision system according to claim 1, wherein The pair and / or more cameras may be housed in a single case, May be separate and the camera is similar to a video camera such as a CCD camera. Digitally converts the data received from the video captured by the camera into digital data. Including a camera with a composite means for converting to   a) at various speeds, with various lights such as infrared (IR), visible light, etc. and insufficient Under any lighting conditions, such as light intensity, the light amplification means Fit;   b) a magnification or reduction device including telescope or microscope means;   c) linear, convex or concave optical image with the desired resolution   The computer vision of claim 1, comprising one or more of the following. system. 3. A computer vision system according to claim 1 or claim 2. What     Detected and / or calculated data, features, definitions, derived The main means of collecting said and / or seen things are absolute, half- And possibly, arranged in the order of one or more hierarchies preset, internal or According to the classification of the external key element, and possibly by said means, Data access immediately, and these   a. Input memory for receiving captured / observed images;   b. Collect all images received by scaling / fixing and / or other shooting Spatial memory for one camera or two cameras for 3D display What The image is stored as a contour image expressed as horizontal and vertical edge coordinates. Updated, constantly updated, spatial memory;   c. A registry in the form of a table for known stored shapes, , Defined and / or given a name and / or separately identified (e.g., Objects, plants, organisms, backgrounds, signs, phenomena, etc.), stored in a special order, known, Images, maps, adjacent above each shape, which can be adjusted for key elements Images and maps, including signatures and / or text data, may be Registry based on shooting distance and shooting distance;   d. One of the images obtained after applying the color filter, movement, movement, and Check at equal time intervals for detection and calculation of speed, etc. Motion detection register;   e. Register known geometric shapes such as squares, steps, etc., preferably in black and white Basic shape register to do;   f. 1. Register for color separation; 2. Preserve the outline of the area and the area before identification. 2. register to live; A register for a table such as an "intrinsic" table, One set of data, definitions, and definitions for conclusions, to draw conclusions 3. The register in which the next table is placed next to or inside the above table; Recurring shape 4. Data table registers for: For key element A database containing:   g. Registers for accessory data: 1. Space counter, con 1. Internal from path etc .; For example, from a camera for enlargement / reduction, for example External things such as those from the user such as exercise speed; additional accessory data;   3. A computer as claimed in claim 1 and claim 2 comprising one or more of the following. -Vision system. 4. The computer vision system according to claim 1, wherein: Thus, the data can be defined as data, According to the requirements and the desired speed and computer language (digital) From images, from calculations, detection, characterization, from definitions and conclusions, From external elements such as printers, speedometers, and / or System accessories such as compass, telescope, distance meter (hereinafter referred to as auxiliary means) And the following or a combination of the following:   a. The direction of the region, or the shape or portion of the region, or the shape (eg, north, south, east, West, up, down, forward, backward, right, left) and coordinate system, or camera, or Area or shape measurements (eg, distance, width, height, depth, size, ratio, and Angle), color (e.g., color percentage level, color variance, , Pot), frame (contour), motion, movement, speed, position, and angle (eg, area, shape) , Depending on the camera);   b. Heat, radiation, voice, smell, and taste characteristics, phenomena, attributes, and detection;   c. Definitions and conclusions (eg temporary names, squares, stairs, objects, cars, Living things, fire, water flow, creatures, quadruped, growth, private riding Right side of the object, change of time);   The computer vision according to any one of claims 1 to 3, wherein System. 5. The computer vision system according to claim 4, wherein:     In unidentified shapes, and / or entire regions / shapes, and / or in envelopes And / or the form in which some movement (eg, movement, movement) occurs Regions, features, and definitions related to shapes and shapes that are visible or partially visible in the image And means for the identification and calculation of said data of conclusions, all of which Unit, its combination is combined with the key element data to operate the system, and The size, distance, scaling, and shooting angle of the shape being viewed Software program to match according to shape size, distance and shooting angle Ram, e-circle, and rules, and multimedia for transfer and 3D display Executed by software programs in the form of media The computer vision system according to claim 4. 6. A computer vision system according to claim 4 or claim 5. What     The data, features, definitions, and conclusions are based on the means and methods for comparison and matching. Registering all or part of the data, depending on the Data / memory and according to one or more predetermined hierarchical order classifications Therefore, the identification of absolute, half, or potential by key elements Performed by the key elements, possibly in advance or from previous pictures For information on known areas, shapes or partial data that are Claims 4 and 5 that can identify a shape or a part thereof. Computer vision system. 7. The computer vision system according to claim 1, wherein: What     Computer, computer component, component, electronic circle, professional One or more of a processor, a computer-based data processing system, etc. Constitute one or more of the steps or a combination thereof; said means are known, new, and compatible. A software program; data obtained from captured images is Said hand to convert to data Step; color identity for the same viewpoint and / or points formed in the captured image Means for matching and definition; distance and distance for identification of features, definitions, and conclusions Means for calculating size and dimensions; size matching and shooting angle match Means for detecting the movement at predetermined time intervals for the form of movement and movement Means for handling "intrinsic" tables; handling recurring shapes Said means for collecting data for key construction; comparison and mapping. Said means for switching; said means for storing various forms of information; Said means for receiving, transferring and withdrawing data; said means comprising power, data protection, 7. A system as claimed in claim 1, including protection of said means, systems and the like. Pewter vision system. 8. A computer vision system according to any one of claims 1 to 7. The stem,     Including calculations, known data, features, and definitions collected for regions, shapes, etc. Accepted and compromised information collected before, during, and after identification may be After a short time, in the form of information, in the usual way and / or in the form of multimedia Transported in a stereoscopic manner and / or in a 3D format, the system may By face means, to users such as robots, devices, blind people, etc. on demand Claims that are automatically stored and / or provided and / or accessible Any of paragraphs 1 to 7 above Computer vision system. 9. A stereoscopic computer vision method, comprising:   a. A pair of identical, aligned, and co-located cameras at a fixed distance M The camera has a shooting angle and an enlargement / reduction arranged on a coordinate axis. For a shadow camera, at a fixed distance M at any shooting distance, from (0: 0) Match the field of view and coordinates of the same site line in both cameras until (Y: X) And form an aligned, optically parallel field of view, the image of the camera Received by the input memory device in the computer, translated into computer language,   b. Based on step A, the pixels of the received image may be captured at the shooting speed and / or At other speeds, by available means, according to coordinates and space counters, Matches each of the previous images in the system's spatial memory registry If there is no match, update the system and enter the two Advance image matching,   c. On the basis of steps A and B, an unidentified area, that is, a certain movement is performed, For an area that is viewed simultaneously by two cameras, the system uses color and other image types. The camera searches for various features and captures them from the start of the line to its matching point. Based on the difference in the number of pixels of each image obtained, the distance of the point in space represented by the pixels Release and Calculate the size, based on which you can calculate the size of the area, distance, etc. ,   d. Based on the steps B and C, the points of the unidentified area and the registered images are mapped. When changing the distance between the camera and the surrounding environment and the distance between the two cameras, Therefore, there are points around the area that do not match, and these depth dots are Create contour lines and / or parting lines in addition to distances approximately equal to For regions, changes in color or darkness between regions and / or changes in region motion / motion The frame or outline of each region is detected and defined by available means. And can   e. Calculation, matching, definition setting, conclusion, depending on the available means One camera to collect additional data that allows the derivation of The video received from the camera is filtered by a color filter and copied for a certain period of time. Enters the motion identification register of the remote and is inspected at regular time intervals. Then, the frame or outline of the detected area is set to the reference shape register and size. And the data from this register allows, for example, It is possible to calculate the size and the movement / movement speed within a given time, additional data Data, part of the data entered, and part of the data previously calculated / detected Can be matched against table registers such as "intrinsic" For example, a stationary object, a vibrating object, or a moving object For example, if the size of the area has changed from one size to another, And other movements and / or movements within the envelope and / or region. For key elements, set definitions, draw conclusions,   f. Derived from ordered data, features, definitions, and conclusions By the key element, the register of the registered shape is provided. The area can be detected, matched and identified by the means   g. Unidentified regions are stored under a temporary name, and identified regions are identified as envelopes and / or Location, position and / or motion and / or motion within the shape In order to always track and monitor such changes, the first Refer to the coordinate data when scanning the dots and additional dots, and 3D computer vision method, saved under the name of the identification area itself. Ten. The computer vision system according to claim 9, wherein the computer vision system The pair and / or more cameras may be housed in a single case, May be separate and the camera is similar to a video camera such as a CCD camera. And the data received from the video captured by the camera. A camera with a combined means for converting the data into digital data, ,   a) at various speeds, with various lights such as infrared (IR), visible light, etc. and insufficient Under any lighting conditions, such as light intensity, the light amplification means Fit;   b) a magnification or reduction device including telescope or microscope means;   c) linear, convex or concave optical image with the desired resolution   10. The computer vision of claim 9, comprising one or more of the following. Method. 11． A computer vision method according to claims 9 and 10. hand,     Detected and / or calculated data, features, definitions, derived The main means of collecting said and / or seen things are absolute, half- And possibly, arranged in the order of one or more hierarchies preset, internal or According to the classification of the external key element, and possibly by said means, Data access immediately, and these   a. Input memory for receiving captured / observed images;   b. Collect all images received by scaling / fixing and / or other shooting Spatial memory for one camera or two cameras for 3D display Thus, the image is represented as horizontal and vertical peripheral coordinates. Spatial memory, stored as a contour image and updated constantly;   c. A registry in the form of a table for known stored shapes, , Defined and / or given a name and / or separately identified (e.g., Objects, plants, organisms, backgrounds, signs, phenomena, etc.), stored in a special order, known, Images, maps, adjacent above each shape, which can be adjusted for key elements Images and maps, including signatures and / or text data, may be Registry based on shooting distance and shooting distance;   d. One of the images obtained after applying the color filter, movement, movement, and Check at equal time intervals for detection and calculation of speed, etc. Motion detection register;   e. Register known geometric shapes such as squares, steps, etc., preferably in black and white Basic shape register to do;   f. 1. Register for color separation; 2. Preserve the outline of the area and the area before identification. 2. register to live; A register for a table such as an "intrinsic" table, One set of data, definitions, and definitions for conclusions, to draw conclusions 3. The register in which the next table is placed next to or inside the above table; Recurring shape 4. Data table registers for: Register for key element; Database   g. Registers for accessory data: 1. Space counter, con 1. Internal from path etc .; For example, from a camera for enlargement / reduction, for example External things such as those from the user such as exercise speed; additional accessory data;   A computer as claimed in claim 9 or claim 10, comprising one or more of the following: Tarvision method. 12． A computer vision method according to claim 9 or claim 10. The data can be defined as data; According to the requirements of the vision and the desired speed and computer language (digital) From images, from calculations, detection, characterization, from definitions and conclusions, space counters From external elements such as printers, speedometers, and / or Accessories for systems such as a pass, a telescope, and a distance meter (hereinafter referred to as auxiliary means) And the following or a combination of the following:   a. The direction of the region, or the shape or portion of the region, or the shape (eg, north, south, east, West, up, down, forward, backward, right, left) and coordinate system, or camera, or Area or shape measurements (eg, distance, width, height, depth, size, ratio, and Angle), color (e.g., color percentage level, color variance, Spot, spot), frame (contour), motion, motion, speed, position, And angle (eg, depending on area, shape, camera);   b. Heat, radiation, voice, smell, and taste characteristics, phenomena, attributes, and detection;   c. Definitions and conclusions (eg temporary names, squares, stairs, objects, cars, Living things, fire, water flow, creatures, quadruped, growth, private riding Right side of the object, change of time);   The computer business according to any one of claims 9 to 10, which is one of the following: Option. 13. A computer vision method according to claim 12, wherein:     In unidentified shapes, and / or entire regions / shapes, and / or in envelopes And / or the form in which some movement (eg, movement, movement) occurs Regions, features, and definitions related to shapes and shapes that are visible or partially visible in the image And means for the identification and calculation of said data of conclusions, all of which Unit, its combination is combined with the key element data to operate the system, and The size, distance, scaling, and shooting angle of the shape being viewed Software program to match according to shape size, distance and shooting angle Ram, e-circle, and rules, and multimedia for transfer and 3D display Claim 12 when executed by a software program in the form of a media Computer vision method according to. 14. The computer vision method according to claim 12 or claim 13. What     The data, features, definitions, and conclusions are based on the means and methods for comparison and matching. Registering all or part of the data, depending on the Data / memory and according to one or more predetermined hierarchical order classifications Therefore, the identification of absolute, half, or potential by key elements Performed by the key elements, possibly in advance or from previous pictures For information on known areas, shapes or partial data that are Claims 12 and 13 wherein the shape, or a part thereof, can be identified. Computer vision method according to. 15． The computer vision method according to claim 9, wherein: hand,     Computer, computer component, component, electronic circle, professional One or more of a processor, a computer-based data processing system, etc. Constitute one or more of the steps or a combination thereof; said means are known, new, and compatible. A software program; data obtained from captured images is Said means for converting to the same viewpoint and / or formed in the captured image. Means for matching and defining color identity for the identified points; Said means for calculating distances and dimensions for identification of features, definitions and conclusions; Means for size matching and matching of shooting angle; form of motion and movement Said means for detecting movements at predetermined time intervals; handling "intrinsic" tables Means for handling recurring shapes; for key construction Said means for collecting data; said means for comparison and matching; Said means for storing information in a form; for receiving, transferring and withdrawing data Said means; said means include power, data protection, protection of said means, systems, etc. A computer vision method according to any one of claims 9 to 14. 16． Computer vision according to any one of claims 9 to 15 The method     Including calculations, known data, features, and definitions collected for regions, shapes, etc. Accepted and stored information collected before, during, and after identification may be immediate or After a short time, in the form of information, in the usual way and / or stereoscopically, the multimedia Transported in media format and / or in 3D format, the system is By face means, to users such as robots, devices, blind people, etc. on demand Claims that are automatically protected and / or provided and / or accessible Item 16. The computer vision method according to any one of Items 9 to 15. 17． Detailed description, reference, description and illustration in the appended claims or accompanying drawings New, implied, stated, or implied herein, or combinations thereof. Computer vision systems and methods, including qualitatively. [Procedure of Amendment] Article 184-8, Paragraph 1 of the Patent Act [Submission date] February 6, 1998 (1998.2.6) [Correction contents]                                  Specification                 Computer stereo vision system and methodField of the invention   The invention is generally viewed in real time by means of a camera, computer, etc. Computer vision and identification, especially with two or more “stereo” cameras Related to D Computer Vision.Background of the Invention   The demand for computer vision is growing in all areas, such as research, astronomy, and robotics. Engineering, industry, agriculture, manufacturing, service, driving, security, assisting blind people, detecting phenomena, etc. It is a daily necessity for solving problems in and for actual use by humans. Today, as both internal and external computer memory capacity grows, The physical dimensions of pewter components are gradually decreasing, and various There are high speed processors. So if you see everything in real time System and method for computer vision that can be uniquely identified If it can be done, everything will be easier and simpler.current situation   Today's computer business using one or two cameras The focus is on individual, defined, known, and primarily static visions of the subject. And their identification is redundant, relatively (image-to-image), partial and focused Was not analyzed and identified at the shooting speed. Also requires many devices such as sensors, lighting, measuring gauges, etc. Is inefficient and does not give a satisfactory solution.Purpose of the invention   It is an object of the present invention to provide a camera connected to a computer, dimensions, movements and other perceptions. Simultaneous and simultaneous capture of all visible shapes by means of computing Providing a system and method that allows analysis and identification at shadow speed It is in.   In addition, the object of the present invention is to provide all automated devices (robots, tools, All tangible objects and phenomena by means of various appropriate visions. It allows you to see in the same way that people can see and identify, Automated devices such as those that communicate in language, Teaches the basic terms of artificial intelligence necessary to communicate with the surrounding environment The movements, tasks and tasks performed by almost everyone, more accurately, efficiently, faster, Good, anywhere, physically difficult to access, dangerous, inaccessible, boring Also to do.   Furthermore, an object of the present invention is to store a real-time 3D image. Enabling broadcast and / or transmission over normal data transfer lines, Thereby, decoding software programs and / or multimedia A, glasses designed for 3-D vision, projection, display, and magnification and / or Special devices allow people to "see" images from a distance It is to be.   It is an object of the present invention to provide equipment, systems, Using circuits and electronic components, basic software, etc. It is to make it possible to configure. And their combination, combination and addition Create and / or extend the device in accordance with the present invention, including its ability to adapt to the environment, by extension, etc. Or can be assembled. Average electronic engineer, computer engineer Near, system analyzers, etc. design, assemble and configure the device in accordance with the present invention. Can be   One of the advantages of the present invention is its simplicity that it can be used with minimal restrictions. The present inventiononlyCan achieve all of the above objects.The present invention   The present invention relates to an innovative system and method for stereoscopic computer vision. Become. 1. Overall, a description of the software that does not form part of the process of the present invention All materials are presented as samples only. Its purpose is to Data vision , And does not reduce the essence of the present invention. 2. Computer vision must be forwarded to serving customers Depending on the data, different types, models and sizes (for example, large, small, medium, Philips, etc.) computer vision systems exist. The explanations given are general, but the basic operating principle in all of them is Are the same. 3. Computer vision is a means of providing "seeing" services. That The assignment looks at sections in space and gathers data, definitions, and terminology for them. And saves the image of the field of view and / or converts the viewed image into code and data, Decrypt as needed, store whatever you need in a register, Transfer of the subject or related data Save 3D data and multimedia by saving at the shooting speed be able to.     If necessary, all should be 3 for the purpose of the computer vision system. -D format or "multimedia" (shape, location, position, size, color of the viewed object , Including data on distance, etc.). 4. Regarding scanning and processing, the two systems and methods are different, but the principle Are identical.     In the image storage system and method, the comparison image is empty. Must be an intermittent image, color, Wr (between two cameras for the same dot) , Etc.), a part of the collected data indicates each dot. The camera image is Scan of the camera image relative to the aerial image, updated by the photo in coordinates Is related to the correlation and depends on the location of the spatial image coordinates compared to the camera image I do.     In this way, the pixel of one camera is adjusted by the device and the coordinates By doing so, in the usual manner, at the shooting speed and / or other speeds, Adjusted for the pixels of the previous image stored in the register All will continue as long as there is a relationship. New movements in the images registered in the video and And / or the shape results in a loss of correspondence and updates, so that the two camera pixels , Pixel to pixel. Data of the same pixel is for each color pixel, Wr difference, etc. Is recorded in the spatial memory. Supports images stored in spatial memory Irrelevant dots are registered as depth lines in the spatial image and the dots in the second image Are subsequently registered separately and matching against spatial memory starts again Is done.     The same applies to the subsequent processing. Continue according to the same principles of detection and comparison of loaded data. But Scanny This is done in different ways to suit the manner of data and data collection. 5. The data collection process is performed in two ways: Wr, coordinate matching, In order to detect movement, correspondence, identification, etc., Comparing the dots of one image with the dots of another image of the same speed and / or Comparing the data with one of the data of two consecutive images; all Eg, frames, geometric lines, geometric shapes based on the data and / or parts thereof Collection / detection of data from data present in the same image, such as shape (frame). Next processing, For artificial intelligence, matching and identification keys, and communication, conclusions, Definitions are derived. 6. If necessary, use part of the Vision Phase for data transfer Can be. For example, in 3D broadcasting, each type of data is represented by an angular momentum. Divide and reduce data structure and transfer bandwidth range and / or the like. Can be. In other places, the code and its data can be To get a 3D image that can be translated / decoded, displayed, projected and enlarged Can be. 7. Computer vision is expected to serve customers, Its working space is usually limited because it is assumed to perform certain operations. Follow Computer vision systems tailored to such customers Must provide the data needed to satisfy the task. 8. The computer vision and / or (preferably) the robot System, customer measurement, hearing, touch (heat, intensity, pressure, wind blowing, etc.), taste, smell Sense, radiation, etc. Additional external assistance may be used, such It may be outside of. Indispensable between customer computer vision and assistance Relationships, compatibility, references, considations, reciprocity, data transfer as appropriate There may be transmissions and the like. 9. The system provides data etc. to the robot and its own and / or robot Can be combined with a separate computing device. Ten. During data collection, basic terms in artificial intelligence are detected. For these basics Words are interpreted by a decoding system and a robot computer to decode images, phenomena, etc. Communication with the environment in which the robot is operated, such as other robots and / or humans It is intended to assist you in conducting a session.List of drawings   FIG. 1 shows a preferred embodiment comprising a possible connection according to the invention with two processors 1 is a schematic column diagram type sketch of a system.   FIG. 2 shows the comparison processing of the data received from the camera and the detected data. Of a preferred system for transferring encoded data to memory It is a low chart.   FIG. 3 shows a plurality of central viewing axes, fields of view of the system shown in FIG. FIG. 2 is a horizontal cross-sectional view schematically illustrating a line and a visual image.   FIG. A / 3 (× 4 magnification) is a parallel visual image of FIG.   FIG. 4 illustrates different distances and different sizes (of the system described in FIG. 1). FIG. 5) is a horizontal cross-sectional view schematically showing the field of view and the shape seen.   FIG. A / 4 (× 4 magnification) shows the system described in FIG. 1 in the shape seen in FIG. 2 shows a view viewed simultaneously by both left and right cameras.Shooting and camera 1. A pair of identical cameras 4 (FIG. 1) are aligned and photographed, including variable and scaling. Turtles whose angles are located on coordinate axes and coexist on a common plane (hereinafter referred to as the horizontal plane) Form an optical parallel field of view (FIG. 3). In other words, each site line The distance to the field is always the same for both cameras from (0: 0) to (Y: X). At a fixed divergence angle distance M (FIG. 3 and A / 3) at all shooting distances is there. The vertical movement of the camera (perpendicular to the horizontal plane) is constant and identical.   The picture taken by the camera is a camera or other equipment 5 (FIG. 1) In the form of digital data translated by the shooting speed (and / or other) Therefore, it is accepted by the input memories P-0 and P-1 (FIG. 1). Alignment, manufacturing, in Stalls and / or objects on any other occasion (preferably by robot) Do it logically. 2. The two or more cameras in one enclosure or separate packaging It is similar to a video camera such as a CD camera. Digital data Including a camera with integrated means for converting to one or more of the following: Including:   a) Various speeds and any light such as IR, visible light, etc., by means such as light amplification Color in any of the lighting conditions, such as, and insufficient Fit for shooting;   b) Enlarging or reducing device including telephoto or microscopic means   c) Optical images containing the desired resolution, linear, concave or convex (FIGS. 3 and A / 3 ).Calculation of distance and dimensions 1. Data, which is the difference between the positions of the corresponding dots in the two camera images on the X axis On the basis of the data Wr and the parallel divergence angle fixed data M [flat at all shooting distances]. Pixel display size can be calculated based on the fixed distance between row lines (FIG. 3). OK: Q = M / Wr 2. The calculation of the distance L between the camera and any dot is calculated for all viewing systems. For the stem, the two cameras must be the same [camera angles must be the same. (In normal conditions, horizontal α⁰, Vertical β⁰)], Physical size (optical image 1) Width), fixed resolution of optical image X: Y, Z enlargement / reduction, M and Wr, and And constants: K0, K1, K2, and K3.   Basic formula     (Α_p) Angle to one pixel, (U) circle of pixel (optical image) Perimeter, (Qm) matching surface¹Size of pixel display in (Qd) shape A And the size of the pixel Xd at d.   1. α_p= Α⁰/ (X * Z)   2. U = Z (X * 360 / α⁰)   3. Qm = 2M   4. Qd = d / Xd   Distance calculation formula   r₀-(Distance) from the radius of the circle to the optical image   r-radius to any point   L—distance from optical image to any point   La—distance to identified shape A (indirect calculation) ¹The coincident plane is defined as the two cameras where the parallel matching dots are the same on the X axis. Refers to the distance when in position.   If Wr = 0, then Q = 2M (the computer knows) and the distance calculation is L = Z (K_Three* M-K₁). 3. By calculating the size of the dot display, the size, length, width, height, surface area, etc. Can be calculated. 4. The order of dots within any frame and / or between adjacent frames, and the X and Y axes The angle can be detected based on the rate of change of the dot position between. Also, geometric lines and shapes, such as straight lines, have a certain periodicity in X and Y changes. , Its signature does not change. The rate of change of the angle of the line is the (tangent / Event) to determine the angle, for example, at each stage of the cycle, the dot is [(+ −1) ) Fixed sign], the angle is -45 °, etc. 5. The calculation of the movement speed V is F (photographing frequency) and T [T is one time duration (Fig. Yes, and that period exists in the computer. ]. 6. Wr also helps to reproduce 3D images in time. 7. Flowchart 101 (FIG. 2) includes factor U (the With the number 5 I chose to be registered next to This is the theoretical rate of change with respect to the change between points (1.25 in the chart).Samples of the system according to the invention and the preferred method of operation   All samples, video clips, flowcharts, etc. are only examples, Whatever the best, errors, inaccuracies, corrections and adjustments to the situation (Like type, translation, etc.). In the process, each processor Performs a given operation, which is an example of these stages, and the processing speed (processor ー), and some changes are made depending on the order of the process steps. It should be emphasized that this is acceptable.Image input process 1. Any received from any camera and translated into computer language (Including color images) are designated as one of the memories P0; P1 (FIG. 1). At line 0, once at line 0, once at line 1, etc. (saving space). Rye Is adjusted to the length of X at the camera resolution. 2. Each code table is given general data, such as a header, some of which are Received from the bot computer, another part of which is It was received from the computer. For example, in a camera, Z-enlarge / reduce, C -The inclination angle of the axis with respect to the horizontal And controlling the camera). Optical image of each image Transfer all four angles (resolution) of the frame dots to the robot's motion coordinates and camera Can be matched. In other words, the basis for subsequent coordinate matching and updating Means that there is data defining its position in space. same Thus, the number of beats t (flow chart (FIG. 2)) is included in the system data. You. 3. In the flowchart (FIG. 2), the PCU-1 (FIG. 1) processor The comparison of the data received from the camera (for example, the pixel a (FIG. 2) with respect to the pixel b) is performed. No, among them, find new data that can be deleted from camera data And register it in the matching table for the code: Q1; Q2; Q3. this Each code is a number in a 4-bit large byte (registered next to it) . The codes are arranged according to their appearance on the Y vertical axis and the X horizontal axis (camera B). Code data in matching tables in the same order and / or other normal order. Be placed.     The PCU-1 processor can be used at any time on the free line to match the shooting speed. Process P-0; P-1 at a rate adjusted accordingly, -0: Insert into PP-1 (FIG. 1) at another position for each imaging cycle.   a. In the flowchart (FIG. 2), square 1— “start” , Square 2- "Read Ya, Yb (read Ya, Yb)", parallelogram 3- " go back. "   b. Circle [101 (FIG. 2)] is the first identical domain in the two camera data. Detect the unit.   c. The circle [102 (FIG. 2)] has a code Q2-color, Q3-region, and Find related data.   Q2: p; Wr; (Y: Xcn); f; (Y: Xx)   Q3: Wr; (Y: Xn); q; u; (Y: Xx)   Where: p is color, Wr is the difference for the same dot in the two cameras, (N, cn) indicates the minimum dot, and (x) indicates the color / line depth / surface. Display the maximum dot, f is the number of dots of the same color, q is the number of dots in one area , U is the number of colors in the area, and (Yn (cn): X); (Y: Xx) is the minimum And the maximum dot.   d. The circle [103 (FIG. 2)] is represented on the X axis by the code Q1-depth line and its associated Data to be detected: When (Y: Xn) is the minimum point, Q1: (p); ( Y: Xn); f (−; +); (Y: Xx), f (−; +) dot on X axis (Negative: Xa camera A, positive: Xb camera B, 3D data), minus sign is " Away ”and the plus sign“ close ”[in artificial intelligence Basic terms], Y: Xx-maximum point (depth lines are drawn because there is no space in the chart Not).   e. The circle [104 (FIG. 2)] indicates the correspondence between the processor and the camera.   f. The circle [105 (FIG. 2)] is used to further process the data. Distribution to various memories for the sake of illustration. 4. All or part of the data of each speed, for example: Q2: (Wr); p; f, Q1: ( p); f (+-) (and shooting pulse) are, for example, television channels, 3D broadcasts, etc. Each data and composition is specified by the code specified to be transmitted / broadcast by By changing the angle of the option and saving the transfer band. In another place Indicates that the transfer has been accepted, separated, translated, and Multiply this by several times, and apply the same to each vertical axis (horizontal axis) Y You can go back the number of times. Therefore, the image can be projected as a 3D image by another method in time. Shadow and / or display.     Similarly, store these data and project them at some other time or place / Can be displayed.Sample of frame point detection method 1. The processor PCU-2 (FIG. 1) includes one of the memories PP0 and PP1. After detecting the data found, this data is complete. Processor P The role of CU-2 is based on encoding and existing encoding data. Frame dots, depth lines, and areas for each color in a fixed order, That is, the headers for the data and the frame dot positions are recorded one after another. Ahead You can find the minimum point, the number of points, and the maximum point on the X-axis from the These data are not available. Therefore, the possible points during the detection of contour points In these data Must be displayed.     The possible contour points are:   1) Yx, f, Xx 2) Yx, f, X   3) Yx, f, Xn (cn), f (q)   4) Y, Xn (cn), f (q)   5) Y, Xx 6) Yn (cn), X   7) Yn (cn), Xx   8) Yn (cn), Xn (cn), f (q) 2. Scanning from left to right, top to bottom, contour identification direction counterclockwise , The processor passes colors, depth lines, and regions through the code table. To the area, in the order of location, or in a form and / or other way. First encountered All codes are processed and the same color, same code and / or adjacent code loop in the image Guo is detected. Then, the provisional header and name of the code (AAn; ABn; ACn). For example, the color is AAn; p; Wr; S, and the depth line is ABn. S, the region is ACn; Wr; S. Each header must be at or within the depth line Regarding the area to be viewed, all the adjacent data including the X-axis and Y-axis data f (− +) There is a column for registering contour dots one after another up to the start point (closed contour). I do. The end point (closure of the contour) is determined by the polar axis [(0: 0) (Y: X)]. In principle, each dot registration identifies adjacent dots It is done after doing. 3. Whenever X is minimal (n; cn), the contour The number f (q) of dots is copied / registered next to it, up to the maximum value on that axis Is done. Whenever Y is the maximum (x), for the same contour, The number is next to the maximum value on that axis (if Y for that X is the smallest, Are registered simultaneously or late). 4. When scanning more lines and / or scanning other lines Must be processed to prevent the same code from being processed twice. Whenever necessary (use identification codes for processing) To prevent this, the code is turned into a continuous code (as a loop), and thus (> > = Change to):     In this way, several processors operate simultaneously on the same database can do. Each processor does its job, and when it's done, Processor continues, etc. Therefore, by the processing according to the size of Wr, two Processors perform processing simultaneously, one from the larger and the other from the smaller. I can. 5. When performing scanning to detect contours for colors, Do not detect adjacent dots of the same color in the code or sequential code database. I have to. Scan line to detect contours for depth lines or regions When performing dot printing, adjacent dots of the same code or continuous code are detected. There must be. 6. When the dot is Xn (cn), the number f (q) of dots on the X axis is represented by S [ S is the total number of dots in the contour, S = S + f (q)], and the code is It changes according to the list of codes and their sequential codes.Contour definition rules 1. Scan the codes according to the previous section and find the data for each code. It is copied to the tentative position DDD. The Xx data of the code is the starting dot of the contour. This data is also the Yx of the contour. Dots are recorded and (no. 1) Is done. The data f of Y is lost and will be completed later. Later, each point in f (q) Pass (obviously), so the dots register one after the other up to the minimum point Xn (cn) [In the first line it is (no. 2)]. Data is copied from DDD And (no. 3). Y is marked, leaving space for that f Then, f (q) on the X axis is added to S [see (c)].     In linear motion (horizontal, vertical), one of the axes (Y: X) remains constant (stationary). And the other proceeds with step (1) with a constant (+ ;-). Detection of adjacent dots Is performed to the right of the axis of motion (at a 90 degree angle) as follows: The axis, which has been stopped and stopped, goes down a straight line by step (1). The dot in the place Move forward until it no longer belongs to the contour (see also a, c, e, g). This point (ring From outside the gull), and thus (see continuations in b, d, fh).     Diagonal movement [adjacent points have the following possibilities: 1) 135 °, 2) 225 °, 3) 315 °, 4) 45 °], the two axes (Y: X) are Step (1) advances each time. Neighboring points on the right side of the direction of motion Points that are further down the diagonal no longer belong to that contour Until then, there is one (45 ° angle) and the other (90 ° angle) [hereinafter, ( i, k, m, o)]. From this point (outside the outline), [(j, l, n, p) reference], For each question mark (?),Where possible and necessary Data must be registered (for that dot and / or the previous dot) ) (> Go to, n = no, y = yes). 2. In the return movement on the line (in the case of a diagonal line, the contour The question must be adjusted, taking into account the record of the On the other hand, once registered backward, twice, code change, the dot of the total (S) Necessary operations regarding addition, minimum value, maximum value, etc. are performed according to the above rules. 3. During scanning, while detecting contours for depth lines or regions, or The code is queried one after another so as not to get out of the line (basically, Xn ( For the movement on cn), the dot on the scanning code, the movement on Xx For dots on continuous codes) and possibly both. 4. The contour of the region may be within or adjacent to the contour of the depth line or at two depths. After detecting the depth line, found between or adjacent to the contours of the stitches, the contour Detect dot table (Internal and external). 5. If it is on a continuous code during contour detection, the next dot will be Even if the dot is not the end dot in the image data (it may be the end dot of the outline ), It must be queried whether it is a starting point within the depth line and within the area within the depth line. No. 6. The rules shown here are examples of contour detection and encompass all details is not. Some of the details are incomplete and must be completed. color Add additional rules if the frame contains additional frames of another color inside There is a need. 7. During this and / or other methods, artificial intelligence data and basic terms Detected for decoding and / or robotic systems to aid in reading and connecting You. For example, the minimum and maximum values are marked for either the X or Y axis of a frame. If not, this indicates a straight line (including the minimum and maximum endpoints). ing. Further, the basic terms are defined based on the data as follows: 1) X Is not defined for the dot for which Yx is defined, so Line "is obtained. 2) X is not defined for dots for which Yn is defined Therefore, a "lower horizontal straight line" is obtained. 3) Y is the dot where Xx is defined. Are not defined, so a "right vertical line" is obtained. 4) Y is Xn The dot for which (cn) is defined is not defined. Is obtained You. 8. At the end of each process and / or in the middle, code tables and data for artificial intelligence , Contours, definitions, and basic terms are M-0; M-1; M-2 notes for each pulse. (Figure 1) and further processed by PCU- (3; 4) processor Is done.PCU- (3; 4) is used for motion / motion in artificial intelligence, geometric line, geometric Detect shapes, basic terms   The PCU-3 processor has completed M-0; M-1; M-2 data completion. Movement / motion between two of the two subsequent memories in a manner that compares them Process and match coordinates: 1. One of the possibilities: the coordinates that entered the first image are the basis for the subsequent coordinates. 2. The second possibility is: according to its movement and the divergence of the camera, The coordinates are calculated, and the photographing data of the camera is temporarily updated. 3. Computer vision equipment, for example, in the first image, a straight east The first point and / or any point on the horizontal line on the plane is zero, and points upward and to the right Is positive, negative in the down and left directions, and each of all pixels at maximum magnification has coordinates A line whose pixels update coordinates from the data of the previous image to the data of the image to be compared. I will be taught to be new. 4. Matching of coordinates and registration in the header are performed using the robot data and camera data. One (small) dynamic identification found in both velocity data according to the movement By matching the contours obtained, the movement / motion This can be performed on the data of the image to be detected. 5. Based on the data collected up to this stage, for motion / motion detection, Between two memories of contour, depth line and color separately, and Y and The comparison between the change in the length of X and the corresponding value in the previous second memory Do the work. 6. Data changes from left to right, from high to low, in the order in which they appear And / or according to the proximity to the camera, ie from the largest Wr in memory Until the end point (smallest Wr) of the contour, it can be detected normally. In addition, Data changes can be detected by more than one processor as needed. You. 7. The PCU-4 processor is defined, for example, to change the distance from the surrounding environment, According to various criteria, such as a uniform direction of motion belonging to the detected geometric shape, and Gather outlines according to the order of proximity to the camera. This contour has not yet been identified (No name in spatial memory MM-0 (FIG. 1)), previous identification (MM-0 The location of the name within is not created by the coordinates of the two contours). If possible, give this contour a definition, for example:   a) adjacent to contours (regions, colors and depth lines) where no movement has taken place The entire contour is identified as an additional contour (including the header) and the temporary name BA n. The significance of artificial intelligence is that it does not move.   b) the adjacent contour where the uniform movement has taken place in size and direction-its loop Guo is detected and registered as an additional contour (including the header) and the temporary name BB n. The significance of artificial intelligence is that it moves.   c) there is a movement inside which is different in size and / or direction from the normal movement. (Or some change has occurred), the outline of BBn is given a temporary name BCn. Can be The significance of artificial intelligence is that it is alive.   d) adjacent contours with a (significant) Wr difference between them and the entire surrounding environment , A tentative name BDn. The significance of artificial intelligence is "floating" That is.       If there is movement within the contour (and within a predetermined range), the temporary name BEn Is given. The significance is "winged bird".   e) if there is time remaining on this processor, and / or another additional If the processor processes the same memory M-0; M-1; M-3 without failure, This The processor can be a complete frame (shape), straight, concave, arcuate (convex), diagonal Contour dots and contours (including angles) and other identifiable geometric lines Detect the definition between Also, geometric shapes such as squares, squares, triangles, circles, ellipses, etc. Shapes are also detected and the basic terms of artificial intelligence are established for these shapes.   f) Lines, contours, etc. whose geometric definitions are not known, or detection takes time Lines, contours, etc. are basic and known to define such a shape / contour definition. The stored basic information for register MM-4 (FIG. 1) of It is matched by matching the size of the contour to the target contour. This Of all registers according to their function and purpose (up to 256 in principle) Compliant with Pewter Vision. 8. During and / or at the end of matching and detection of movements, various contours, etc. In addition, the decoding system and / or the robot can provide many Know additional basic terms for:   a. Length Y remained constant and moved up (changed location), meaning-"up Rose! "   b. Length Y remained constant and moved down (changed location), meaning-"down I fell down. "   c. Length Y is smaller, meaning-"shorter"   d. Length X is smaller, meaning-"smaller"   e. Significance of length Y increased-"increased"   f. Significance of X and Y increased-"expanded"   g. Significance of X and Y becoming smaller-"contracted" 9. During processing and / or at other times, many basic terms in artificial intelligence Can be detected / derived. This is the shape and / or Is used to decode phenomena and / or terms and / or communications in many ways. Useful. Ten. At the end of each process and / or during the process, code, data, contour, definition, and A table of the basic terms of artificial intelligence is stored in the memory MM- according to its attributes, locations, and coordinates. 0 (spatial memory) and additional processor PCU- (5; 6 ...) ( It is further processed and identified by FIG. 1). 11． The spatial memory MM-0 stores, for example, a horizontal line of 180 ° and a vertical line of 120 °. Certain spaces and / or all spatial areas are covered as needed.Shape identification Key, identification   Data, contours, features, phenomena, definitions, results collected during matching and detection Etc. are stored according to their attributes and locations in the spatial memory MM-0. A special order adapted to the same order in which the shape data is arranged, And for matching and identification (data vs. data) and / or Is the key factor according to the "intrinsic" table, etc. To form   Since the unidentified area (contour) is very small at each shooting speed, the In the margins. Unidentified shape data arranged in a special order can be used as a key factor. Constitute. In this way, all or part of the means and frequency, "intrinsic" tables etc. , Like words in a dictionary, the classification is language, and the order of key elements is the order of letters , If the whole word is the key, compare the data with other images / records, and / or Or compare the data with known stored shape data and image / record Can be detected, matched, and identified. This saved shape data is the same Are arranged in a special order, according to the key elements, and "reliable", "half and half" , "Reasonable", "possible". 1. The identification is based on the data of the register MM-0 in the shape seen and the stored constant data. This is performed between data of a register MM-1 (FIG. 1) having a known shape. Dimension data In the MM-0, the matching of the size, angle, and shooting distance is based on Or the size of the dimension data of the stored shape according to the scaling data And the photographic / recording angle standard. 2. Shape data stored in MM-1 is adjusted according to the matching key Display the shape on computer vision according to the given size and distance Thus, the data of each shape can be obtained by decoding and / or the data Conditions can also be defined by logical analysis.     The shapes are in a specific order (e.g., color, geometric, etc.) that matches the decryption key ( Dimensions), defined and arranged according to "Sure", "Half", "Reasonable" and "Possible" There is a possibility. " 3. There is an exercise memory MM-2 (FIG. 1), which is used to calculate the exercise / motion speed. For example, data on all changes in the “live” shape can be Over a defined period of time, such as only a desired distance and / or to a desired distance , Every pulse, every 10 pulses, every second, etc. 4. In this or another memory, the total movement / motion in each pulse It is possible to save all changes and pulses. Therefore, 3D data and / or Or it is possible to store multimedia, so that translation, decoding Display and / or projection, etc., at different times and in different places even after zooming and magnifying. It is possible to 5. It is also possible to provide a pseudo memory MM-3 (FIG. 1). This memo Lee contains the outline of the space in which the robot 7 (FIG. 1) functions, and inside the outline of the space. Displays the outlines of important shapes and spaces, where some identifying details are provided. Details, such as name (add one or two data and / or one or two words) ) And the like. If the robot is somewhere At any time, the memory can know the location, and if the location changes, Know that the location of the guo, shape, etc. has changed (what kind of change and where) be able to. So Then, the pseudo space is updated according to the new reality and the same registration method. External data A PM memory (FIG. 1) may be provided to receive data.wrap up 1. Computer vision systems can be any kind of daze, including lasers Light, flash, etc., wherever the system can be Shall be protected as much as possible from physical damage to 2. Computer vision systems are compatible and require user input Can be operated according to 3. The computer vision system includes software program 6 (FIG. 1), Electronic and general circles, distance and / or scaling data that make the system work , And according to the size of the dimensional data of the stored shape and the standard of shooting / recording A software program for adjusting the size of the viewed shape and the shooting angle, and Transfer requested data to external for transmission, 3D display and multimedia Any other software programs and any other software programs required. Program. 4. Received images, code, data, calculations, images in spatial memory, and For the received images or maps or data stored in all memory of the system Other information that may include data, including code and data only, and / or input or space Including stereoscopic images, to users such as robots, Depending on the requirements and design, it may be sent directly by any method, The user can extract data when needed according to the system design. 5. All of the above operations are performed by the processor and matching, detecting, processing, and calculating Is performed by additional processing means. These are computer vision Equipment, components, one or more electronic circuits necessary and adapted for the purpose of Or a combination thereof.     These roles are assigned by each part, part, electronic circle, system, etc. It plays its role without interfering with the Dination and compatible. For example, the processor No. 1 in operation A Since his part has been completed, the processor No. 2 in complex operation A Parts, electronic circles, etc., indicate that the part in charge of their own may be continued. The processor No. 2 is notified.Commercial implementation 1. Computer vision can be used as follows.   a. Viewer (monitor, analyze, decode, report, transfer Etc.)   b. Collect data in any way and store it in a database or register As viewers (complete, partial, classified by any form, etc.) 2. Due to the many possibilities of computer vision, some standards Computer vision systems can be used, and each standard system Adapted to provide certain services and perform certain tasks, and thus "for protection" Can be used for On the other hand, for special requirements, A computer vision system. 3. Each computer vision system has standard and special abilities. You. 4. The computer vision system is designed by a designer or a constituent at the design stage System to be part of the user Or as a separate unit serving the user.                                The scope of the claims 1. A stereoscopic computer vision system, comprising:   a. A pair of identical, aligned, and co-located cameras at a fixed distance M The camera has a shooting angle and an enlargement / reduction arranged on a coordinate axis. For a shadow camera, at a fixed distance M at any shooting distance, from (0: 0) Match the field of view and coordinates of the same site line in both cameras until (Y: X) And form an aligned, optically parallel field of view, the image of the camera Received by the input memory device in the computer, translated into computer language,   b. Based on step a, the pixels of the received image may be processed in various ways For example: The pixels of the image are matched one by one by the device and the data Data collected / detected for headers, including pulse / speed, scaling / color, For encoding, the distance Wr of the same dot between the two cameras is equal For location, number of adjacent horizontal dots of the same color and / or distance, or for 3D 1. depth line, etc .; The pixels of one camera follow the shooting speed and / or the other speed According to the order, the pixels of the previous image stored in the spatial memory register And then by matching the coordinates, the new movement is matched by the device as well as/ Or, if there is no match in shape, update and pixels in both images are mapped to each other. And the data of the same pixel is spatially recorded with respect to the pixel of each color, Wr difference, etc. Unmatched pixels are recorded as depth line dots in the aerial image. Recorded, the pixels of the other images are recorded separately, related, etc.,   c. Size, dimensions, distance, etc., based on the device and data Wr based on step b And the device and data share the work between the processors. To help you play 3D images in time,   d. Based on the previous steps, the color, depth line, and outline of the area, as well as adjacent, And / or near, etc. regions having uniform motion, groups of depth lines, These are added to the existing data as a data table, and the vertical line Size is detected and the contour size and angle are matched with the stored basic shape. Geometric lines and / or geometric shapes are detected, such as by Through comparison, straight line, circle, triangle, concave, immobile, floating, living etc. , Definitions for artificial intelligence have been derived, while measurements, hearing, taste, smell, touch, etc. External auxiliary terms to help identify and / or communicate with the surrounding environment And   e. Based on the previous process, the data of the continuous image The same static that is compared by the device and / or the contours are found at both speeds Stop shape , The coordinates are matched to the continuous data, thus the robot The imaging deflection angle that allows the camera to balance is detected and the Detects horizontal / vertical movement / motion and calculates degree of change, direction, angle, etc. The coordinates can be matched by the change data stored at a predetermined interval speed. And the speed of the movement / motion can be calculated and a part of the detected data Can be stored in various ways, and changes in speed can be stored one after another. 3D data for projection / display in some way after translation / decoding and enlargement Data can be stored, broadcasted, transferred, etc., in any way, anytime, anywhere. Can,   f. Based on the previous steps, some or all of the data collected by the device For each data, according to its attributes, coordinate matching, spatial memory Unidentified shape data that can be stored and arranged in a specific order "Intrinsic", comprising all elements and by some or all devices and by incidence Image according to table etc. and recorded image / record / data size, angle and distance / Record / Data as "Sure," "Half," "Reasonable," and "Possible." Arranged in the same special order, adjusted to the key elements according to the stored Matching / detection / identification of image / record / data of known shape It is possible,   g. During detection, unidentified contours and their data are recorded under a temporary name and identified. The saved shape is recorded under your own name, and the artificial memory is displayed with the contour name display and Operate the robot in the form of contours with basic shapes and some identification data Can be added to the surrounding environment, sometimes from a robot and / or vision system. 3D computer vision system updated by the system. 2. The computer vision system according to claim 1, wherein The pair and / or more cameras may be housed in a single case, May be separate and the camera is similar to a video camera such as a CCD camera. Converts the data received from the video captured by the camera to digital data A camera comprising a combined means for performing   a) at various speeds, with various lights such as infrared (IR), visible light, etc. and insufficient Under any lighting conditions, such as light intensity, the light amplification means Fit;   b) a magnification or reduction device including telescope or microscope means;   c) linear, convex or concave optical image with the desired resolution   The computer vision of claim 1, comprising one or more of the following. system. 3. A computer vision system according to claim 1 or claim 2. What     Detected and compared data that can be defined as data And / or computerized, defined, derived, viewed, etc. Information (eg code table, size code, color, contour, geometric line, geometric shape Collect, store, state, definitions, and / or common terms, and / or within artificial intelligence) Main device for the size and angle of the stored data. Match coordinates and separation according to key elements according to matching of size and angle To identify the preserved shape in terms of dimensions Therefore, a device that allows access to data, wherein the key element is “ One, two, three, four, five, etc. In a hierarchical order above, which may be internal or external, the device comprises:   a. Input memory for receiving camera / viewed images;   b. Spatial notes for one or two cameras and / or data for 3D And integrates with image preservation methods by matching coordinates Separately or in addition to other data in another memory In the meantime, for the second image, the data of both cameras All, including fixed, scaled records, etc. All images, data and tables are integrated and updated for some part of the defined space. New and adjusted on every pulse, images and data are stored in horizontal and vertical perimeters. Stored as indicated and constantly updated;   c. A register in the form of a table for a known stored shape, defined Any shape that may be given and / or given a name and / or separately identified ( (E.g. screwdriver, finger, wall, Dear Sir, sign, phenomenon, etc.) Images, maps, signs, etc. that are adjacent to, above, and connected to Data and tables), data, codes, etc. Registered in a specific known order, adjusted for Adjusted according to the shooting / recording standards according to the method and the size of the shooting angle;   d. A register for detecting movement at various speeds. The transformation is stored and adjusted to a defined, known time interval, the extent and rate of change And, based on the number of shooting frequencies, the size and speed of the exercise can be calculated;   e. Code, data, and all changes in horizontal and vertical dimensions at a given speed Registers to save one after another, after translation / decryption and expansion, another time And at another location, allowing the image to be displayed / projected again;   f. Registers for contours of basic shapes that cannot be defined and detected in a reasonable amount of time Where, with respect to the detected contour, Can be detected, thus deriving the definition as a supplement to the key data And;   g. Collect and handle tables as needed, their data on contours, tables, And its data and tables of terms, trigonometric and other definitions, of artificial intelligence Memory and registers for encoding basic terms;   h. Ancillary data received from devices such as robots, audio devices, and detection devices Register for;   i. In order for a computer vision device to perform its function, Computer detection and ratio to perform all operations required for Registers for software with capacity for comparison, handling, and identification;   A computer web according to claims 1 and 2, comprising one of the following. John system. 4. A method for stereoscopic computer vision, the method comprising:   a. A pair of identical, aligned, and co-located cameras at a fixed distance M The camera has a shooting angle and an enlargement / reduction arranged on a coordinate axis. Shadow turtle Due to the constant distance M at any photographing distance, (0: 0) to (Y: X) align and coordinate the field of view and coordinates of the same sight line in both cameras until An aligned, optically parallel field of view is formed and the images of the camera are stored in a computer. Received by the input memory device, translated into computer language,   b. Based on step a, the pixels of the received image may be processed in various ways For example: The pixels of the image are matched one by one by the device and the data Data collected / detected for headers, including pulse / speed, scaling / color, For encoding, the distance Wr of the same dot between the two cameras is equal For location, number of adjacent horizontal dots of the same color and / or distance, or for 3D 1. depth line, etc .; The pixels of one camera follow the shooting speed and / or the other speed According to the order, the pixels of the previous image stored in the spatial memory register And then by matching the coordinates, the new movement is matched by the device And / or if there is no match in shape, update the pixels in both images , And the data of the same pixel is represented by the spatial Pixels that do not match are recorded in the spatial memory, And the pixels of the other images are recorded separately, related, etc.   c. Based on the process b, the device and the data Wr Sizes, dimensions, distances, etc. can be calculated, and the equipment and data Work to distribute 3D images in time,   d. Based on the previous steps, the color, depth line, and outline of the area, as well as adjacent, And / or near, etc. regions having uniform motion, groups of depth lines, These are added to the existing data as a data table, and the vertical line Size is detected and the contour size and angle are matched with the stored basic shape. Geometric lines and / or geometric shapes are detected, such as by Through comparison, straight line, circle, triangle, concave, immobile, floating, living etc. , Definitions for artificial intelligence have been derived, while measurements, hearing, taste, smell, touch, etc. External auxiliary terms to help identify and / or communicate with the surrounding environment And   e. Based on the previous process, the data of the continuous image The same static that is compared by the device and / or the contours are found at both speeds By matching to the stop shape, the coordinates are matched to the continuous data, thus The imaging deflection angle that allows the robot to balance is detected, and various contours are detected. Horizontal / vertical movements / motions are detected and the degree of change, direction, angle, etc. is measured Matches coordinates based on calculated and saved change data at specified intervals By doing, the speed of the movement / motion can be calculated, Some of the detected data can be stored in various ways, and changes in speed It can be stored one after another, and after being translated / decrypted and expanded, Save anytime, anywhere as 3D data for shadow / display in various ways, Broadcast, transfer, etc.   f. Based on the previous steps, some or all of the data collected by the device For each data, according to its attributes, coordinate matching, spatial memory Unidentified shape data that can be stored and arranged in a specific order "Intrinsic", comprising all elements and by some or all devices and by incidence Image according to table etc. and recorded image / record / data size, angle and distance / Record / Data as "Sure," "Half," "Reasonable," and "Possible." Arranged in the same special order, adjusted to the key elements according to the stored Matching / detection / identification of image / record / data of known shape It is possible,   g. During detection, unidentified contours and their data are recorded under a temporary name and identified. The saved shape is recorded under your own name, and the artificial memory is displayed with the contour name display and Operate the robot in the form of contours with basic shapes and some identification data Can be added to the surrounding environment, sometimes from a robot and / or vision system. Updated by the system, for stereoscopic computer vision the method of. 5. A method for computer vision according to claim 4, wherein:   The pair and / or more cameras may be housed in a single case, Or a separate camera, similar to a video camera such as a CCD camera Digitally converts the data received from the video captured by the camera into digital data. Including a camera with a composite means for converting to   a) at various speeds, with various lights such as infrared (IR), visible light, etc. and insufficient Under any lighting conditions, such as light intensity, the light amplification means Fit;   b) a magnification or reduction device including telescope or microscope means;   c) linear, convex or concave optical image with the desired resolution   5. The computer vision of claim 4, comprising one or more of the following. Way for. 6. A method for computer vision according to claims 4 and 5. And     Detected and compared data that can be defined as data And / or computerized, defined, derived, viewed, etc. Information (eg code tables, size codes, colors, contours, geometric lines, geometric Collect, store, shape, definitions, and / or common terms, and / or within artificial intelligence) Main device for the size and angle of the stored data, Match coordinates and separation according to key elements, according to size and angle matching To identify the stored shape in terms of dimensions by Device that allows access to data, wherein the key element is Pre-established ones such as "N", "Half", "Reasonable", and "Possible" In the above hierarchical order, which may be internal or external, said device comprises:   a. Input memory for receiving camera / viewed images;   b. Spatial notes for one or two cameras and / or data for 3D And integrates with image preservation methods by matching coordinates Separately or in addition to other data in another memory In the meantime, for the second image, the data of both cameras All images and data obtained by fixed, scaled recording and others, including And tables are integrated and updated for some part of the defined space, and all Adjusted on a per-loose basis, images and data are shown in horizontal and vertical peripheral coordinates Saved in and constantly updated;   c. A register in the form of a table for a known stored shape, defined And / or given a name And / or any separately identified shape (eg, screwdriver, (Finger, wall, Dear Sir, sign, phenomenon, etc.) Images, maps, signs (including data and tables related to them), data Data, codes, etc., in a specific, known order adjusted for key elements. The registered and dimension data depends on the shape, its dimensions, and the size of the shooting angle. Adjusted according to shooting / recording standards;   d. A register for detecting movement at various speeds. The transformation is stored and adjusted to a defined, known time interval, the extent and rate of change And, based on the number of shooting frequencies, the size and speed of the exercise can be calculated;   e. Code, data, and all changes in horizontal and vertical dimensions at a given speed Registers to save one after another, after translation / decryption and expansion, another time And at another location, allowing the image to be displayed / projected again;   f. Registers for contours of basic shapes that cannot be defined and detected in a reasonable amount of time Where, with respect to the detected contour, Can be detected, thus deriving the definition as a supplement to the key data And;   g. Collect and handle tables and outlines as necessary Its data, tables, and terms for that data and tables, trigonometrically and otherwise And memory for encoding the definitions of terms, basic terms of artificial intelligence, etc. Jister;   h. Ancillary data received from devices such as robots, audio devices, and detection devices Register for;   i. In order for a computer vision device to perform its function, Computer detection and ratio to perform all operations required for Registers for software with capacity for comparison, handling, and identification;   A computer web according to claims 4 and 5, comprising one of the following. The way for John. 7. A computer vision system according to any one of claims 1 to 6. A stem and a method,     Computers, computer components, components, electronic circuits, processors Server, computerized data processing system, etc. Constitute one or more of the steps or a combination thereof, said means being capable of doing all that is necessary. A new, compatible, known software program with the ability to For converting data obtained from a captured image into digital data; Means: color, code, feature, identity, point of view and / or adjacent point and / or And / or formed in the image of the camera Normal form and / or artificial intelligence with respect to contours stored in some registers Said means for matching, comparing, detecting and defining; codes, data Tables, contours, terms, features, definitions, calculations, basic terms in artificial intelligence, etc. for data Means for creating and detecting tables for distances; dimensions, coordinates, angles, speeds Said means for calculation, comparison, and adjustment of movements / movements at predetermined time intervals Said means for detecting movements of the loop; for handling frequent and / or "intrinsic" tables Said means for collecting data for key elements; constant and defined Compare registers according to a specified, known order and / or according to keys Said means for switching, said means for preserving various forms of data; Said means for receiving, transferring and withdrawing data; said means for input, data, Any one of claims 1 to 6, including protection, said means, systems, methods, etc. And a computer vision system and method. 8. A computer vision system according to any one of claims 1 to 7. A stem and a method,     Well-known and trivial calculations, data, features, common definitions, and conclusions Collection before, during, and after identification, including basic terminology and / or artificial intelligence. The input and stored information that was collected, and the information collected about the area, shape, etc. Immediately or after a short time, data information, speed In the usual manner and / or in 3D and / or multimedia form The system is transferred by means of a formula, and the system is On demand and / or automatically for users such as robots, equipment, blind people, etc. Claim 1 to claim 1 and / or provided and / or accessible. A computer vision system and method according to any of claims 7 to 13. 9. Detailed description, reference, description and illustration in the appended claims or accompanying drawings New, implied, stated, or implied herein, or combinations thereof. Computer vision systems and methods, including qualitatively. FIG.FIG. 2FIG. 3FIG. 4

────────────────────────────────────────────────── ─── Continuation of front page    (81) Designated countries EP (AT, BE, CH, DE, DK, ES, FI, FR, GB, GR, IE, IT, L U, MC, NL, PT, SE), OA (BF, BJ, CF) , CG, CI, CM, GA, GN, ML, MR, NE, SN, TD, TG), AP (KE, LS, MW, SD, S Z, UG), UA (AM, AZ, BY, KG, KZ, MD , RU, TJ, TM), AL, AM, AU, BA, BB , BG, BR, CA, CN, CU, CZ, EE, FI, GE, HU, IL, IS, JP, KG, KP, KR, L C, LK, LR, LT, LV, MD, MG, MK, MN , MX, NO, NZ, PL, RO, SG, SI, SK, TR, TT, UA, US, UZ, VN [Continuation of summary] Includes data table registers such as tables. This register Can detect data, features, and definitions, and The purpose is to draw conclusions for the statement. Recognized by the data, features, definitions and conclusions Unidentified shape compatible with stored shape registers Key can be configured for the unidentified shape Is completely and almost instantaneously identified.

Claims (1)

[Claims] 1. A stereoscopic computer vision system, comprising:   a. A pair of identical, aligned, and co-located cameras at a fixed distance M The camera has a shooting angle and an enlargement / reduction arranged on a coordinate axis. For a shadow camera, at a fixed distance M at any shooting distance, from (0: 0) Match the field of view and coordinates of the same site line in both cameras until (Y: X) And form an aligned, optically parallel field of view, the image of the camera Received by the input memory device in the computer, translated into computer language,   b. Based on step A, the pixels of the received image may be captured at the shooting speed and / or At other speeds, by available means, according to coordinates and space counters, Matches each of the previous images in the system's spatial memory registry If there is no match, update the system and enter the two Advance image matching,   c. On the basis of steps A and B, an unidentified area, that is, a certain movement is performed, For an area that is viewed simultaneously by two cameras, the system uses color and other image types. The camera searches for various features and captures them from the start of the line to its matching point. Based on the difference in the number of pixels Then, the distance is calculated, and the calculated distance indicates the sky represented by the pixel. Data such as the size of points in the interval can be calculated, and based on this, the area Size and distance can be calculated,   d. Based on the steps B and C, the points of the unidentified area and the registered images are mapped. When changing the distance between the camera and the surrounding environment and the distance between the two cameras, Therefore, there are points around the area that do not match, and these depth dots are Create contour lines and / or parting lines in addition to distances approximately equal to For regions, changes in color or darkness between regions and / or changes in region motion / motion The frame or outline of each region is detected and defined by available means. And can   e. Calculation, matching, definition setting, conclusion, depending on the available means One camera to collect additional data that allows the derivation of The video received from the camera is filtered by a color filter and copied for a certain period of time. Enters the motion identification register of the remote and is inspected at regular time intervals. Then, the frame or outline of the detected area is set to the reference shape register and size. And the data from this register allows, for example, It is possible to calculate the size and the movement / movement speed within a given time, additional data Data, part of the data entered, and Depending on the part of the data previously calculated / detected, the register of the table, e.g. Can be matched, for example, Object, moving object, and, for example, the size of an area Movement and / or movement in size and in the envelope and / or region Set definitions and draw conclusions for other key elements, such as whether Start out   f. Derived from ordered data, features, definitions, and conclusions By the key element, the register of the registered shape is provided. The area can be detected, matched and identified by the means   g. Unidentified regions are stored under a temporary name, and identified regions are identified as envelopes and / or Location, position and / or motion and / or motion within the shape In order to always track and monitor such changes, the first Refer to the coordinate data when scanning the dots and additional dots, and 3D computer vision system stored under the name of the identification area itself. 2. The computer vision system according to claim 1, wherein The pair and / or more cameras may be housed in a single case, Is separate Often, the camera is similar to a video camera such as a CCD camera, A complex to convert the data received from the video captured by La to digital data A camera comprising means,   a) at various speeds, with various lights such as infrared (IR), visible light, etc. and insufficient Under any lighting conditions, such as light intensity, the light amplification means Fit;   b) a magnification or reduction device including telescope or microscope means;   c) linear, convex or concave optical image with the desired resolution   The computer vision of claim 1, comprising one or more of the following. system. 3. A computer vision system according to claim 1 or claim 2. What     Detected and / or calculated data, features, definitions, derived The main means of collecting said and / or seen things are absolute, half- And possibly, arranged in the order of one or more hierarchies preset, internal or According to the classification of the external key element, and possibly by said means, Data access immediately, and these   a. Input memory for receiving captured / observed images;   b. Collect all images received by scaling / fixing and / or other shooting One camera or 3D display Spatial memory for two cameras, the images being stored in horizontal and vertical directions. Sky, stored and constantly updated as a contour image, expressed as perimeter coordinates Interim memory;   c. A registry in the form of a table for known stored shapes, , Defined and / or given a name and / or separately identified (e.g., Objects, plants, organisms, backgrounds, signs, phenomena, etc.), stored in a special order, known, Images, maps, adjacent above each shape, which can be adjusted for key elements Images and maps, including signatures and / or text data, may be Registry based on shooting distance and shooting distance;   d. One of the images obtained after applying the color filter, movement, movement, and Check at equal time intervals for detection and calculation of speed, etc. Motion detection register;   e. Register known geometric shapes such as squares, steps, etc., preferably in black and white Basic shape register to do;   f. 1. Register for color separation; 2. Preserve the outline of the area and the area before identification. 2. register to live; A register for a table such as an "intrinsic" table, One set of data, definitions, and definitions for conclusions, to draw conclusions 3. The register in which the next table is placed next to or inside the above table; Recurring shape For 4. data table registers; A register for the key element; Database;   g. Register for accessory data; Space counter, con 1. Internal from path etc .; For example, from a camera for enlargement / reduction, for example External things such as those from the user such as exercise speed; additional accessory data;   3. A computer as claimed in claim 1 and claim 2 comprising one or more of the following. -Vision system. 4. The computer vision system according to claim 1, wherein: Thus, the data can be defined as data, According to the requirements and the desired speed and computer language (digital) From images, from calculations, detection, characterization, from definitions and conclusions, From external elements such as printers, speedometers, and / or System accessories such as compass, telescope, distance meter (hereinafter referred to as auxiliary means) And the following or a combination of the following:   a. The direction of the region, or the shape or portion of the region, or the shape (eg, north, south, east, West, up, down, forward, backward, right, left) and coordinate system, or camera, or Area or shape measurements (eg, distance, width, height, depth, size, ratio, and Angle), color (for example, color Level, color dispersion, scattered dots, spots), frame (outline), dynamic Work, motion, speed, position, and angle (eg, by area, shape, camera);   b. Heat, radiation, voice, smell, and taste characteristics, phenomena, attributes, and detection;   c. Definitions and conclusions (eg temporary names, squares, stairs, objects, cars, Living things, fire, water flow, creatures, quadruped, growth, private riding Right side of the object, change of time);   The computer vision according to any one of claims 1 to 3, wherein System. 5. The computer vision system according to claim 4, wherein:     In unidentified shapes, and / or entire regions / shapes, and / or in envelopes And / or the form in which some movement (eg, movement, movement) occurs Regions, features, and definitions related to shapes and shapes that are visible or partially visible in the image And means for the identification and calculation of said data of conclusions, all of which Unit, its combination is combined with the key element data to operate the system, and The size, distance, scaling, and shooting angle of the shape being viewed Software program to match according to shape size, distance and shooting angle Ram, e-circle, and rules, and multimedia for transfer and 3D display Claim 4 when executed by a software program in the form of a media The computer vision system as described. 6. A computer vision system according to claim 4 or claim 5. What     The data, features, definitions, and conclusions are based on the means and methods for comparison and matching. Registering all or part of the data, depending on the Data / memory and according to one or more predetermined hierarchical order classifications Therefore, the identification of absolute, half, or potential by key elements Performed by the key elements, possibly in advance or from previous pictures For information on known areas, shapes or partial data that are Claims 4 and 5 that can identify a shape or a part thereof. Computer vision system. 7. The computer vision system according to claim 1, wherein: What     Computer, computer component, component, electronic circle, professional One or more of a processor, a computer-based data processing system, etc. Constitute one or more of the steps or a combination thereof; said means are known, new, and compatible. A software program; captured image Said means for converting data obtained from the same into digital data; Or matching and definition of color identity for points formed in the captured image Means for calculating distances and dimensions for identification of features, definitions, and conclusions Means for size matching and matching of shooting angle; motion Said means for detecting movement at predetermined time intervals for the type of movement; Means for handling tables; means for handling recurring shapes; keys Said means for collecting data for construction; said means for comparison and matching Means for storing information in various forms; receiving and transferring data; Said means for withdrawing; said means for power, data protection, said means, system, etc. 7. The computer vision system according to claim 1, which includes protection of the computer vision system. Stem. 8. A computer vision system according to any one of claims 1 to 7. The stem,     Including calculations, known data, features, and definitions collected for regions, shapes, etc. Accepted and compromised information collected before, during, and after identification may be After a short time, in the form of information, in the usual way and / or in the form of multimedia Transported in a stereoscopic manner and / or in a 3D format, the system may By face means, to users such as robots, devices, blind people, etc. on demand And / or automatically stored and / or provided and / or accessible The computer business according to any one of claims 1 to 7, wherein the computer business System. 9. A stereoscopic computer vision method, comprising:   a. A pair of identical, aligned, and co-located cameras at a fixed distance M The camera has a shooting angle and an enlargement / reduction arranged on a coordinate axis. For a shadow camera, at a fixed distance M at any shooting distance, from (0: 0) Match the field of view and coordinates of the same site line in both cameras until (Y: X) And form an aligned, optically parallel field of view, the image of the camera Received by the input memory device in the computer, translated into computer language,   b. Based on step A, the pixels of the received image may be captured at the shooting speed and / or At other speeds, by available means, according to coordinates and space counters, Matches each of the previous images in the system's spatial memory registry If there is no match, update the system and enter the two Advance image matching,   c. On the basis of steps A and B, an unidentified area, that is, a certain movement is performed, For an area that is viewed simultaneously by two cameras, the system uses color and other image types. The camera searches for various features and captures them from the start of the line to its matching point. Based on the difference in the number of pixels Then, the distance is calculated, and the calculated distance indicates the sky represented by the pixel. Data such as the size of points in the interval can be calculated, and based on this, the area Size and distance can be calculated,   d. Based on the steps B and C, the points of the unidentified area and the registered images are mapped. When changing the distance between the camera and the surrounding environment and the distance between the two cameras, Therefore, there are points around the area that do not match, and these depth dots are Create contour lines and / or parting lines in addition to distances approximately equal to For regions, changes in color or darkness between regions and / or changes in region motion / motion The frame or outline of each region is detected and defined by available means. And can   e. Calculation, matching, definition setting, conclusion, depending on the available means One camera to collect additional data that allows the derivation of The video received from the camera is filtered by a color filter and copied for a certain period of time. Enters the motion identification register of the remote and is inspected at regular time intervals. Then, the frame or outline of the detected area is set to the reference shape register and size. And the data from this register allows, for example, It is possible to calculate the size and the movement / movement speed within a given time, additional data Data, part of the data entered, and Depending on the part of the data previously calculated / detected, the register of the table, e.g. Can be matched, for example, Object, moving object, and, for example, the size of an area Movement and / or movement in size and in the envelope and / or region Set definitions and draw conclusions for other key elements, such as whether Start out   f. Derived from ordered data, features, definitions, and conclusions By the key element, the register of the registered shape is provided. The area can be detected, matched and identified by the means   g. Unidentified regions are stored under a temporary name, and identified regions are identified as envelopes and / or Location, position and / or motion and / or motion within the shape In order to always track and monitor such changes, the first Refer to the coordinate data when scanning the dots and additional dots, and 3D computer vision method, saved under the name of the identification area itself. Ten. The computer vision system according to claim 9, wherein the computer vision system The pair and / or more cameras may be housed in a single case, Is separate Often, the camera is similar to a video camera such as a CCD camera, A complex to convert the data received from the video captured by La to digital data A camera comprising means,   a) at various speeds, with various lights such as infrared (IR), visible light, etc. and insufficient Under any lighting conditions, such as light intensity, the light amplification means Fit;   b) a magnification or reduction device including telescope or microscope means;   c) linear, convex or concave optical image with the desired resolution   10. The computer vision of claim 9, comprising one or more of the following. Method. 11． A computer vision method according to claims 9 and 10. hand,     Detected and / or calculated data, features, definitions, derived The main means of collecting said and / or seen things are absolute, half- And possibly, arranged in the order of one or more hierarchies preset, internal or According to the classification of the external key element, and possibly by said means, Data access immediately, and these   a. Input memory for receiving captured / observed images;   b. Collect all images received by scaling / fixing and / or other shooting One camera or 3D display Spatial memory for two cameras, the images being stored in horizontal and vertical directions. Sky, stored and constantly updated as a contour image, expressed as perimeter coordinates Interim memory;   c. A registry in the form of a table for known stored shapes, , Defined and / or given a name and / or separately identified (e.g., Objects, plants, organisms, backgrounds, signs, phenomena, etc.), stored in a special order, known, Images, maps, adjacent above each shape, which can be adjusted for key elements Images and maps, including signatures and / or text data, may be Registry based on shooting distance and shooting distance;   d. One of the images obtained after applying the color filter, movement, movement, and Check at equal time intervals for detection and calculation of speed, etc. Motion detection register;   e. Register known geometric shapes such as squares, steps, etc., preferably in black and white Basic shape register to do;   f. 1. Register for color separation; 2. Preserve the outline of the area and the area before identification. 2. register to live; A register for a table such as an "intrinsic" table, One set of data, definitions, and definitions for conclusions, to draw conclusions 3. The register in which the next table is placed next to or inside the above table; Recurring shape For 4. data table registers; A register for the key element; Database;   g. Registers for accessory data: 1. Space counter, con 1. Internal from path etc .; For example, from a camera for enlargement / reduction, for example External things such as those from the user such as exercise speed; additional accessory data;   A computer as claimed in claim 9 or claim 10, comprising one or more of the following: Tarvision method. 12． A computer vision method according to claim 9 or claim 10. The data can be defined as data; According to the requirements of the vision and the desired speed and computer language (digital) From images, from calculations, detection, characterization, from definitions and conclusions, space counters From external elements such as printers, speedometers, and / or Accessories for systems such as a pass, a telescope, and a distance meter (hereinafter referred to as auxiliary means) And the following or a combination of the following:   a. The direction of the region, or the shape or portion of the region, or the shape (eg, north, south, east, West, up, down, forward, backward, right, left) and coordinate system, or camera, or Area or shape measurements (eg, distance, width, height, depth, size, ratio, and Angle), color (for example, color Level, color dispersion, scattered dots, spots), frame (outline), dynamic Work, motion, speed, position, and angle (eg, by area, shape, camera);   b. Heat, radiation, voice, smell, and taste characteristics, phenomena, attributes, and detection;   c. Definitions and conclusions (eg temporary names, squares, stairs, objects, cars, Living things, fire, water flow, creatures, quadruped, growth, private riding Right side of the object, change of time);   The computer business according to any one of claims 9 to 10, which is one of the following: Option. 13. A computer vision method according to claim 12, wherein:     In unidentified shapes, and / or entire regions / shapes, and / or in envelopes And / or the form in which some movement (eg, movement, movement) occurs Regions, features, and definitions related to shapes and shapes that are visible or partially visible in the image And means for the identification and calculation of said data of conclusions, all of which Unit, its combination is combined with the key element data to operate the system, and The size, distance, scaling, and shooting angle of the shape being viewed Software program to match according to shape size, distance and shooting angle Ram, e-circle, and rules, and multimedia for transfer and 3D display Claim 12 when executed by a software program in the form of a media Computer vision method according to. 14. The computer vision method according to claim 12 or claim 13. What     The data, features, definitions, and conclusions are based on the means and methods for comparison and matching. Registering all or part of the data, depending on the Data / memory and according to one or more predetermined hierarchical order classifications Therefore, the identification of absolute, half, or potential by key elements Performed by the key elements, possibly in advance or from previous pictures For information on known areas, shapes or partial data that are Claims 12 and 13 wherein the shape, or a part thereof, can be identified. Computer vision method according to. 15． The computer vision method according to claim 9, wherein: hand,     Computer, computer component, component, electronic circle, professional One or more of a processor, a computer-based data processing system, etc. Constitute one or more of the steps or a combination thereof; said means are known, new, and compatible. A software program; data obtained from captured images is Said hand to convert to data Step; color identity for the same viewpoint and / or points formed in the captured image Means for matching and definition; distance and distance for identification of features, definitions, and conclusions Means for calculating size and dimensions; size matching and shooting angle match Means for detecting the movement at predetermined time intervals for the form of movement and movement Means for handling "intrinsic" tables; handling recurring shapes Said means for collecting data for key construction; comparison and mapping. Said means for switching; said means for storing various forms of information; Said means for receiving, transferring and withdrawing data; said means comprising power, data protection, The method according to claims 9 to 14, which includes protection of said means, system and the like. Computer vision method. 16． Computer vision according to any one of claims 9 to 15 The method     Including calculations, known data, features, and definitions collected for regions, shapes, etc. Accepted and stored information collected before, during, and after identification may be immediate or After a short time, in the form of information, in the usual way and / or stereoscopically, the multimedia Transported in media format and / or in 3D format, the system is By face means, to users such as robots, devices, blind people, etc. on demand Claims that are automatically protected and / or provided and / or accessible Any of paragraphs 9 to 15 Computer vision method. 17． Detailed description, reference, description and illustration in the appended claims or accompanying drawings New, implied, stated, or implied herein, or combinations thereof. Computer vision systems and methods, including qualitatively.