41results about How to "Image processing is simple" patented technology

The invention provides an active vision non-contact servo mechanism parameter measuring device, consisting of a measuring screen, a servo mechanism mounting platform, an image acquiring and processing system, and a red light laser light source, wherein a servo mechanism to be detected is provided with the red light laser light source which is positioned on the servo mechanism mounting platform and is arranged in front of the measuring screen, a digital camera is arranged behind the measuring screen to image on the screen. Firstly, the device shoots a picture of a gridiron pattern on the measuring screen by the camera which is demarcated; secondly, the servo mechanism controls a main frame to give a command so as to control the rotation of the servo mechanism to be detected, a computer extracts laser point coordinates of the image, the rotational angle at each moment of the servo mechanism to be detected is calculated and recorded in real time, the motion parameters of the servo mechanism to be detected such as the angular velocity, the pointing accuracy, and the overshoot are calculated by the rotational angle; the measuring method is simple, convenient and flexible, and the device and the measuring method realize the measurement in real time and with high accuracy of the movement parameters of a biaxial servo mechanism through the non-contact way.

The invention relates to a non-contact complex surface cone volume measurement system and method. The method includes the steps: firstly, providing three linear lasers to form a light sectioning plane and marking the contour of the section of a cone to be measured; secondly, providing a vertical lifter, moving the light sectioning plane relative to the cone to be measured; thirdly, providing a black and white camera for acquiring images and transmitting acquired image information to a computer for processing the image information to finally obtain the volume and a three-dimensional diagram of the cone to be measured. The measuring system and the method are simple and practicable, high in scanning speed and precision and low in requirement for the cone and environments and low in system cost.

The invention discloses a scene depth measurement method, equipment and an imaging device. The scene depth measurement method comprises the following steps: S1, driving a varifocal lens unit to acquire at least two images from the same scene under at least two focal powers, wherein the magnifying powers of the images are the same; S2, respectively acquiring a relative defocusing degree value or focusing degree value of each image; S3, acquiring the relative depth distribution of the scene through a DFD algorithm or a DFF algorithm. Through the adoption of the scene depth measurement method, the equipment and the imaging device disclosed by the invention, the operation of changing the magnifying powers of the images of the same scene is unnecessary, the image processing is simplified, and the relative depth distribution of the scene can be rapidly and effectively measured.

The invention provides a detection device for surface-shape errors of a solar trench type curved surface reflector, relating to the field of mirror surface detection, which solves the problems that the when the surface-shape errors of the trench type curved surface reflector are detected by the conventional detection device, the detection process is complicated and time-consuming and is not suitable for rapid and online detection, and the subsequent image processing is complicated. The device comprises a light source device, a self-calibration device and an image acquisition and processing device, wherein the light source device comprises a laser, a beam expander and a beam splitter group; light beams emitted from the laser are expanded by the beam expander, are split by the beam splitter group and are reflected; a test light beam split by the beam splitter group is projected on the curved surface reflector to be detected; a self-calibration light beam reflected by the beam splitter group is transmitted into the self-calibration device; the self-calibration device is used for calibrating the parallel degree of the test light beam projected to the curved surface reflector to be detected; and the image acquisition and processing device is used receiving light spot information reflected by the curved surface reflector to be detected and determining the surface-shape errors of the curved surface reflector to be detected. The detection device is easy to operate and has low cost.

The invention discloses a double-flexible-cantilever-beam vibration measurement and control device and method based on laser binocular vision. The device comprises two flexible cantilever beams connected in series. The flexible cantilever beams are provided with piezoelectric ceramic drivers, piezoelectric ceramic sensors and accelerometers. The device also comprises a line laser, an industrial camera, a servo motor and a planetary speed reducer. The line laser generates a laser light resource, and combined with the industrial camera, in a complete non-contact mode, the bending vibration of the cantilever beams is detected. The speed of the servo motor is reduced by the planetary speed reducer, and the servo motor is used for controlling the rotation of the cantilever beams, receiving a vibration suppression signal obtained by a vibration control algorithm and suppressing the vibration of the cantilever beams at the same time. The device can be used for the vibration measurement and control of a space flexible joint, laser is used as the light resource without the affection of sun light, thus the working stability is maintained, and a reference is provided for the vibration measurement and vibration suppression of a flexible aerospace structure.

An imaging device includes: a plurality of pixels (22) which output a pixel signal obtained by photoelectrically converting an incident light; and pixel mixing units (25, 26, 27, 28) which mix pixel signals from pixels of adjacent rows. The pixel mixing units (25, 26, 27, 28) perform a first pixel mixing operation which mixes pixel signals in the same columns and a second pixel mixing operation which mixes pixel signals in the same rows.

The invention relates to an unmanned aerial vehicle. The unmanned aerial vehicle comprises an unmanned aerial vehicle main assembly and a camera assembly, and the camera assembly is detachably fixed on the unmanned aerial vehicle main assembly; the camera assembly comprises a camera main body, a top lens and a bottom lens; the top lens is fixed on one end of the camera main body, and the bottom lens is fixed on one end of the camera main body opposite to the top lens; the angle of view of the top lens and the angle of view of the bottom lens are added to form a total visual angle greater thanor equal to 360 degrees; and the camera assembly is fixed on a appropriate portion of the unmanned aerial vehicle main assembly so that the fields of views of the top lens and the bottom lens are notobstructed by the unmanned aerial vehicle main assembly. The above-mentioned unmanned aerial vehicle can simplify the process of image-processing and reduce hardware costs.

It is possible to linearly arrange cells and indicate particular data in the order of colors of respective sells. A code system uses an optical symbol decoding method capable of reading the color arrangement if the arrangement continuity and the linear shape are maintained. The data expression method may be other than 'the order of colors'. For example, it is possible to employ a method for allocating a numeric value to each of the colors by one-to-one correspondence (R - 0, B = 1, etc.), a method for allocating data by a color transition ('CM' = 'MY' = 'YC' = 0, 'CY' = 'YM' = 'MC' = 1, etc.), a method for allocating data for a combination of colors, and the like.

A DLS model-based high-definition processing method and system for cadre personnel files are disclosed, an image processing DLS model is constructed to record and analyze the image boundary information. The square of L2 norm of the gradient is used as a regular constraint condition to constrain the gradient of the boundary. The regularization denoising algorithm and the image smoothing distribution logic are introduced to minimize the total variation while obtaining the best image. Based on the image processing DLS model, the cadre personnel files are scanned to obtain scanned images and stored, and meanwhile, the scanned images are transmitted to the server for high-definition processing to obtain high-definition images. The invention innovates the image high-definition processing technology, and avoids the problems of image color, shape distortion, discrimination reduction and the like caused by the unreasonable application of the image acquisition technology.

This invention discloses an optical in situ stress monitoring method for epitaxial growth. The method comprises: dividing a laser beam into at least two parallel laser beams, casting to the surface of an epitaxial growth sheet, reflecting, transmitting via optical elements, imaging on an image collecting apparatus, processing, and acquiring the surface stress information. The wavelength of the laser beams is smaller than that corresponding to the bandwidth of the epitaxial growth sheet material, which can eliminate system error, and largely improve the precision of stress measurement.

The invention relates to a 3D camera for robot disordered sorting and a measurement method thereof, and the method comprises the steps: employing an active speckle projection technology, and projecting a designed speckle pattern to the surface of a measured object through a projection unit; then, starting the two cameras to synchronously acquire an image once; and finally, performing matching calculation and three-dimensional analysis on the acquired image by using a digital image correlation method, thereby realizing rapid reconstruction of the three-dimensional point cloud of the measured surface. Only one image needs to be collected in each measurement, the image processing is simpler, and the measurement speed is greatly improved. The technical problems that a traditional structured light 3D camera is small in measurement breadth and low in calculation speed are solved. And the projection breadth of the projection unit is larger, the brightness is higher, and the measurement requirement of 1.5 m large visual field of logistics sorting can be met.

The invention discloses an illumination region control device and control method for an LED phototherapy device. The control device comprises a host computer connected with an imaging CCD and a microprojector. The light output port of the micro projector is provided with an optical switch array, An LED array for outputting therapeutic light is disposed directly in front of the optical switch array. The optical switches in the optical switch array and the LED lights in the LED array have the same number and corresponding positions. One end of each optical switch is connected in series with thecorresponding LED light, and the other end of each optical switch is connected to ground. One end of each LED light is connected in series with the corresponding optical switch, and the other end ofeach LED light is connected to the level. After obtaining the image of a focus area, the control device converts the image into a highlighted image and projects the highlighted image onto the opticalswitch array, and the optical switch array turns on the LED array light source so that a therapeutic light source having consistent shape of the light-emitting area and highlighted image is formed, thereby achieving the control of a precise phototherapy to illumination area control.

The present invention discloses a prepositive image automatic segmentation system used for a lingual surface instrument. The prepositive image automatic segmentation system comprises an FPGA controller and a PC, the signal end of the FPGA controller is connected with a data acquisition module, and the control end is also connected with an image processor. A sensor set is arranged inside the imageprocessor, the output end of the sensor set is connected with an image pre-processing module, and the output end of the image pre-processing module is connected with the PC via a data extraction module. A NiosII soft core processing module is arranged in the PC, and a template matcher is arranged in the NiosII soft core processing module and receives a data signal of the data extraction module. Atongue picture is segmented via an image automatic segmentation algorithm and is transmitted to the PC to process, and the PC is utilized to process the image, compares and adjusts the size of the image automatically to completely match and block a lingual surface, so that the image processing is simpler, the image information interference is less, the speed is fast, and the accuracy is high.

The present invention discloses a movie theatre attendance statistics method. Character patterns are arranged at the backrest on each seat in a movie theatre, an infrared camera is arranged at the upper portion of the movie theatre, the images of seats including characters are collected and transmitted to an image processing module, and characters are extracted through adoption of the character identification technology to obtain vacant seat information. According to the method, the general seat human being is subjected to fuzzy identification and is changed to the identification of standard characters, and hard things make simple. The patterns of characters can be directly printed on the backrests of the seats with no need for taking spaces accessing power supply and maintaining. The technology of identifying and extracting standard characters from a video image is perfect and reliable with no need for distinguishing the overlapping portions of a plurality of pictures, the image processing is simplified, and the system is stable and reliable and the cost performance is improved. Through adoption of an infrared or visible light camera, the movie theatre attendance statistics method can be widely applied to other occasions requiring the statistics of the number of seats such as a classroom, a conference hall, a library, a waiting lounge, a bank queue machine and the like.

The present disclosure provides a sensing device, a tactile sensing device, and a tactile sensor. The sensing device includes: a photosensitive sensor; a pinhole array structure located on the photosensitive sensor; an elastomer layer positioned on one side, far away from the photosensitive sensor, of the pinhole array structure; a reflecting layer positioned on one side, far away from the pinholearray structure, of the elastomer layer; and a light source located between the reflective layer and the pinhole array structure.

The invention discloses a method for positioning a workpiece in a three-dimensional space based on machine vision, and the method comprises the steps: carrying out the shooting of a to-be-detected workpiece through a camera, and obtaining an original image of the workpiece in a detection region; performing image preprocessing on the original image; carrying out binaryzation in the preprocessing process based on k entropy in the generalized statistical mechanics theory; and obtaining position information of the target workpiece in a pixel coordinate system through ROI region extraction and image edge detection. The object distance of the image can be obtained by analyzing the pinhole camera model, then the real-time position information of the target workpiece under the robot base coordinate system can be obtained through coordinate transformation, and accurate positioning of the target workpiece is achieved.

The invention discloses a shaft hole centering guide method in a workpiece assembly process based on machine vision. The method comprises the steps that a to-be-detected workpiece is shot through an industrial camera, and an image of the workpiece in a detection region is obtained; and a 3d shape model for matching is created from a 3d object model of the to-be-detected workpiece. A virtual camera is used for shooting the 3d object model in a specified attitude range, and the 3d shape model is generated through images at different visual angles. A greedy algorithm is adopted to retrieve in a specified image pyramid range, the most suitable matching object is found, the posture of the matching object is obtained, and a tail end of a robot is adjusted to perform correct shaft hole centering guide operation through posture analysis.