44results about How to "Facilitate image recognition" patented technology

The invention discloses a multifunctional intelligent blind guiding method. The multifunctional intelligent blind guiding method comprises the following steps of receiving a voice command, and navigating; obtaining image data of an obstacle; carrying out recognition processing on digital image data, judging that threat is formed by the obstacle if the recognized obstacle is a static body, generating an ultrasonic ranging control command, and generating and outputting voice alarm information or vibration prompt information according to a ranging result if the ranging result is within a preset alarm range; and generating an ultrasonic ranging control command if the recognized obstacle is a dynamic body, carrying out threat anticipation calculation according to a ranging result, and generating and outputting voice alarm information or vibration prompt information if the obstacle is within a preset alarm range. The invention also discloses a multifunctional intelligent blind guiding device and a processor. According to the multifunctional intelligent blind guiding method, the multifunctional intelligent blind guiding device and the processor, which are disclosed by the invention, the blind guiding service can be more safely, conveniently and reliably provided for the blind, the blind is prevented from walking more distance and meeting the obstacle, and thus the blind can quickly and safely arrive at the destination.

The semiconductor device has a simplified structure which includes a package structure in which a member for mounting a semiconductor element is separate from a member including a signal input/output unit. A microphone package is configured with a cover and a substrate. A microphone chip and a circuit element are adhered and fixed to a top surface of a recess formed in the cover. A plurality of bonding pads are arranged on the lower surface of the cover on the outer side of the recess. A bonding wire is connected to the circuit element and the bonding pad. The substrate includes a signal input/output terminal serving as the signal input/output unit, and a connection electrode, conducted with the signal input/output terminal, is arranged facing the bonding pad on the upper surface of the substrate. The substrate, cover, connection electrode, and bonding pad are joined with a conductive member.

The embodiment of the invention discloses a face recognition method and related equipment, which are used for improving the performance of a face recognition system. The method provided by the embodiment of the invention comprises the following steps: obtaining an environment illumination condition; Determining an image acquisition target mode according to the environment illumination condition, wherein the target mode comprises a visible light mode and/or an infrared induction mode; Collecting image information in the target mode; Performing face recognition processing on the image information to obtain target face feature information; And matching a user corresponding to the target face feature information in a database. The target mode comprises a visible light mode and/or an infrared induction mode, so that compared with a single image recognition technology in the prior art, the method can flexibly use the visible light mode and/or the infrared induction mode to obtain image information, is more beneficial to image recognition and face feature extraction, and can improve the performance of a face recognition system.

The invention provides a visible light and X-ray image matching method and a visible light and X-ray image matching device based on a conveyor belt identifier in order to solve the problem of matching of object images shot by visible light and object images scanned by X-ray in a security check field scene. A network high-definition camera module is installed above an inlet of the X-ray security inspection machine, and digital identification strips made of metal materials are installed on the two sides of the conveying belt. The method comprises: using a deep focusing convolutional residual network to carry out article detection on a visible light snapshot image and an X-ray scanning image, and using a convolutional neural network using the residual network as a model to carry out automatic identification on digits on the images; and calculating the number on the marking strip corresponding to the article through projection by using the detected article coordinates on the visible light snapshot image and the X-ray scanning image; and then realizing matching of the visible light and the X-ray article image through matching of digits on the visible light snapshot image and digits on the X-ray scanning image, so that a security inspector can confirm the visible light article image corresponding to the X-ray article image more conveniently, and security inspection efficiency is improved.

The invention discloses a video monitoring system that includes: an image preprocessing unit that is used for converting image data come from a camera into a binary code stream; an image compression unit that processes compression encoding to the binary code stream with RGB format; an encryption unit that is used for encrypting image data in compression encoding process; a center processor that is used for scheduling controlling whole system. The compression process used by the image compression unit is JPEG or JPEG-2000 standard. The image of system encrypting prevents the image from being tampering, generates image that can be identified after being tampered, has uniqueness, can regard the image data generated directly as only accuracy proof and is used for various security protection system; inter-frame information in the compression process is not be compressed which ensures high definition of image and is convenient for identifying later image.

A semiconductor device of the present invention includes a lead frame having an island portion having a roughened upper surface and side faces, and an unroughened lower surface, and also having a plurality of leads having roughened inner lead portions and unroughened outer lead portions; a semiconductor chip placed on the upper surface of the island portion of the lead frame; a plurality of electrode pads provided on the upper surface of the semiconductor chip; a plurality of wires connecting the plurality of electrode pads and the plurality of leads; and a resin molding the semiconductor chip.

The invention discloses a sub-eye optimal arrangement method of a compound eye system based on three-dimensional detection. A regular polyhedron is subjected to subdivision with the frequency being v,subdivided vertexes are used as sub-eye centers for distribution structure design, and the method has the main advantages that subdivision with the frequency being v subdivides each face of the regular icosahedron for v times, and the sub-eye arrangement uniformity is greatly improved compared with an original hexagonal structure; the compound eyeball shell designed and processed by using the regular polyhedron subdivision method can realize more dense arrangement. Manufacturing is convenient, and subsequent processing such as image recognition, splicing and multi-channel fusion is also facilitated. The defect that regular hexagons are seriously deformed due to the fact that hexagons cannot be arranged on spherical surfaces in a sealed connection mode in a traditional arrangement method is overcome.

The invention discloses an automatic shield segment assembling method based on machine vision, belongs to the technical field of automatic shield construction segment assembling, and relates to the automatic shield segment assembling method based on the machine vision. The method includes the steps that firstly, an industrial camera is installed and calibrated, and a target ArUco code is arranged on a segment; then, four feature angular points of the target ArUco code are detected, and an expected position point of the segment to be assembled is predicted; a deviation between a to-be-assembled point and the expected position point is calculated, and an error value between current speed and expected speed of a translation, lifting and rotation mechanism is calculated; and the expected speed of each mechanism of a segment erector is obtained by adopting a visual servo strategy based on an image, and finally the control quantity of each execution mechanism is obtained through joint speed closed-loop control, so that the automatic assembly of the segments is realized. According to the automatic shield segment assembling method based on the machine vision, a machine vision technology is introduced to replace a manual control handle to complete segment splicing, and the efficiency is high. The real-time prediction of a pose of the segment is indirectly realized based on a target, and the assembly precision and the assembly efficiency of the segment are effectively improved.

Disclosed is a lead frame for a semiconductor device (10), which comprises a plurality of terminals (13) having one end portion sealed with a resin. A portion (16) of each terminal (13), which is to be sealed with a resin, is composed of a pentagonal or higher polygonal prism, or an irregular prism having at least one notch or groove which is formed on the peripheral surface and extends in the vertical direction. The portion (16) to be sealed with a resin is formed by etching or pressing, and an exposed portion (18) serving as the lower half of each terminal (13) is formed by etching.

The invention provides a crop monitoring device for intelligent agricultural planting. The device comprises a combined box, wherein a through receding groove is formed in the inner wall of the obverse side of the combined box, the bottom surface of the combined box is fixedly connected with a base box, a main controller is fixedly installed on the obverse side of the base box, and a capturing mechanism is arranged on the inner side of the top of the combined box. According to the crop monitoring device, the combined box, the receding groove, the capturing mechanism and a luring mechanism cooperate with one another and are controlled by the main controller in a unified mode, agricultural pests are lured into the combined box, then striking plates which suddenly rotate at a high speed and are located inside the capturing mechanism are utilized to collide with the agricultural pests, and the agricultural pests are dizzy or are killed to fall to a combined bearing plate, so that the early-period collection capture and killing operation of the agricultural pests is achieved. Compared with two existing agricultural pest monitoring devices, the crop monitoring device has the advantages that the pests are complete and convenient to identify, a specific collecting capturing and killing structure is not affected by the use time and the environment, and a novel operation mode is created.

The invention relates to a method for adjusting parameters of a hearing system (1) comprising a hearing aid device (3) and a communication device (5) separate from the hearing aid device, for acoustically coupling the hearing aid device to a hearing aid At least one component (9, 10) of the wearer's ear is mounted on the hearing device, wherein at least one image of the or each component for the acoustic coupling is recorded by means of a camera (14) of the communication device, wherein according to the recorded The or each image identifies a product identity of the or each component, wherein the pre-identified or each component is associated with an acoustic transfer characteristic curve (35), and wherein parameters of the hearing system are adjusted based on the acoustic transfer characteristic curve. The invention also relates to a hearing system comprising a hearing device with at least one component for acoustically coupling the hearing device to an ear of a hearing aid wearer, wherein parameters of the hearing system can be adjusted based on the acoustic transfer characteristic.

The invention relates to the technical field of transportation management, in particular to an intelligent transportation management system and method based on image recognition. The system comprises:a cargo analysis subsystem which is used for obtaining an image A of a vehicle contour with a reference mark, obtaining a preset reference length of the reference mark, and obtaining the mark lengthof the reference mark according to the image A and the reference length, and is also used for acquiring theoretical cargo carrying information and calculating the actual cargo carrying volume according to the marked length and the theoretical cargo carrying information; and an auxiliary analysis subsystem which is used for acquiring the standard auxiliary distance, the actual auxiliary distance and the loading material, and generating an auxiliary analysis conclusion according to the standard auxiliary distance, the actual auxiliary distance, the loading material and the actual loading volume.By the adoption of the scheme, effective statistics can be conducted on the actual loading volume of the vehicle at a time according to the reference marks, the actual loading volume is checked through downward movement of the carriage, and therefore the situation that another building material is hidden under the building material is avoided, and the transportation situation is effectively supervised.

A mirror, when attached along a curved surface, facilitates recognition of an image in a blind spot by reducing a feeling of strangeness caused by the aspect ratio of the image. Many linearly-extending strip reflective surfaces 5a are arranged on a surface of a sheet 2 to be parallel and adjacent to one another. An inclination angle A formed between a width direction of each of the strip reflective surfaces 5a and a surface direction of the sheet 2 is sequentially changed with an increase in a distance in a direction transverse to an extension direction of the strip reflective surfaces 5a. An aggregate of the strip reflective surfaces 5a has a cylindrical convex mirror function, and the mirror 1 is bendable in a cylindrical shape about an axis parallel to the direction transverse to the extension direction of the strip reflective surfaces 5a.

The invention discloses a detection method for a coal ash fusion tester. The detection method comprises the following steps: regulating the relative positions of a tray, an index plate, a video cameraand a detector to enable a corresponding ash cone to continue to rotate by a preset angle to reach the vision preset position of the video camera when the detector detects the passing of each index hole; controlling the tray and the index plate to synchronously rotate uniformly, vertically arranging a rotating axis, and enabling the rotating axis to coincide with a central axis of the index plate; controlling the video camera to start continuously acquiring images each time the detector detects the passing of one index hole, controlling the video camera to stop acquiring the images after theash cone reaches the vision preset position, controlling the video camera to starting continuously acquiring images again until the detector detects the passing the next index hole, and controlling the video camera to stop acquiring the image after the ash cone reaches the vision preset position. The ash cone in the image is guaranteed to be located on the vision preset position of the video camera, the shooting accuracy is improved, the image recognition is facilitated, and the maintenance times are reduced.

The invention relates to the technical field of image processing, and discloses an image processing method and system, so as to improve the precision and efficiency of image restoration processing. The method comprises the following steps: constructing a convolutional neural network model to realize color spot identification of a face image and marking a corresponding color spot area; using a camera located right ahead the middle of the face for collecting a figure image, recognizing a corresponding face area, calling the convolutional neural network model to judge whether the corresponding face image has color spots or not, and if yes, marking the corresponding color spot area for a user to further confirm whether the area is an area to be repaired or not; in the image processing coordinate system, preferably performing transplanting restoration through the color of the symmetrical region, and ensuring the restored color to be consistent with the natural color; and when the conditionsare not matched, color restoration processing is carried out according to the pixel values corresponding to the adjacent grids without color mutation, and ensuring that the restored color is smoothlyand naturally matched with the periphery.