7results about How to "Implement defect detection" patented technology

This invention provides a fully automatic sock inspection and sorting machine, including a frame. A sock-supporting rotating assembly is mounted on the frame to open the heel of the sock and rotate it 360° horizontally. A detection component is mounted on the side of the sock-supporting rotating assembly. The detection component includes a detection module for acquiring image information of the sock heel at a first height and a first tilt angle, and at a second height and a second tilt angle, for defect detection. This invention utilizes the sock-supporting rotating assembly to open the heel of the sock, while the detection component can adjust its height and tilt angle, thereby achieving defect detection of both the sock leg and the heel, further improving detection accuracy. Furthermore, by dividing the sock into upper and lower parts for inspection, the influence of sock length on the detection results is reduced.

The utility model provides a scraper defect detection device which comprises a machine table, a product movement module and two sets of visual detection mechanisms, the two sets of visual detection mechanisms are arranged on the two sides of the product movement module respectively, and each set of visual detection mechanism comprises an edge detection camera, a side face detection camera, an edge detection light source and a side face detection light source. The product movement module comprises a conveying linear movement module and a product platform, the product platform is used for fixing a scraper to be detected, the central axis of the edge detection camera and the plane of the product platform form an inclined included angle, and the central axis of the side face detection camera is parallel to the plane of the product platform. The automatic feeding and double-face detection device has the automatic feeding and double-face detection functions, the two sides of the product movement module are each provided with a visual detection mechanism, the surface of a scraper body and the edge of a blade of the scraper can be more accurately detected while comprehensive covering of the scraper is achieved, and then defect detection of the scraper can be efficiently and accurately achieved.

The application discloses a kind of based on binocular vision's electric wire insulator defect detection method and system, first, the multi-view image of insulator is acquired, and the preliminary three-dimensional point cloud data of insulator is calculated according to multi-view image;Then, preliminary three-dimensional point cloud data is carried out point cloud denoising and dynamic light compensation processing, and the three-dimensional point cloud data after optimization is obtained;Then, calling three-dimensional defect detection model integrated with multiple modules such as multi-scale feature extraction module, mixed attention module, feature alignment module, feature enhancement module, detects according to three-dimensional point cloud data and the two-dimensional image of insulator, and the defect detection result of insulator is obtained.The three-dimensional image of the application embodiment is obtained by binocular vision camera, and the shooting angle problem is solved, and the change of illumination is handled using dynamic light compensation, and the three-dimensional defect detection model is used to overcome the bottleneck of micro-defect detection, effectively improve the accuracy of defect detection, so as to realize the high-precision insulator defect detection.

The utility model discloses a geomembrane defect multidirectional detection rod under a heavy cover condition, which comprises a vertical telescopic rod, a handheld short rod and a mounting rod, a plurality of groups of rotating sleeves are arranged on the mounting rod, and horizontal telescopic rods are symmetrically fixed on two sides of each rotating sleeve; a steel chisel clamp is fixed to the far end of each horizontal telescopic rod, a steel chisel is fixed to each steel chisel clamp, and each steel chisel is connected with a detection line. The multi-directional detection rod comprises the vertical telescopic rod and the horizontal telescopic rods which can be telescopically stored, so that the size of the multi-directional detection rod is reduced, the steel chisels are stored and fixed in cooperation with the steel chisel buckles, the size of the multi-directional detection rod is further reduced, and the multi-directional detection rod is convenient to carry and use. The multidirectional detection rod comprises four groups of steel chisels which can be stored, two steel chisels in each group form a group of receiving electrodes M and receiving electrodes N. The four groups of receiving electrodes M and receiving electrodes N can realize electric signal detection in eight directions at a detection position.

This application discloses an electrode tab imaging device and an electrode die-cutting system, belonging to the field of electrode die-cutting technology. The electrode tab imaging device includes: a first guide roller; a second guide roller disposed below the first guide roller, forming a vertically extending electrode feeding channel between the second guide roller and the first guide roller; and an imaging device disposed between the first guide roller and the second guide roller, which images the electrode tabs when the electrode tabs on both sides of the electrode are between the first guide roller and the second guide roller. According to the electrode tab imaging device of this application, by using the first guide roller and the second guide roller to make the electrode feed vertically, and then using the imaging device to image the vertical electrode between the first guide roller and the second guide roller, imaging is performed when the electrode tabs are vertical, avoiding contact between the electrode tabs and the roller surface. This enables the detection of pinhole defects and eliminates the ranging deviation caused by scratches on the roller surface. It also avoids the ranging deviation caused by electrode tab sagging, resulting in more comprehensive and accurate defect detection.

The invention discloses an optical lens precision detection device, and belongs to the technical field of lens detection. Comprising a detection table, and a placement hole is formed in the inner side of the detection table; the supporting assembly is arranged at the bottom of the detection table corresponding to the placing hole and is used for horizontally supporting the to-be-detected lens; the annular light source assembly is arranged at the top of the detection table corresponding to the placing hole and is used for performing dark field illumination on the to-be-detected lens at a low incident angle; according to the invention, the uniformly arranged piezoelectric ceramic actuators are adopted to support the to-be-detected lens, and the flexible plunger heads at the tops of the piezoelectric ceramic actuators are subjected to micron-order adjustment according to the pressure detection signals, so that the pressure values of a plurality of fulcrums are equal, and the feedback adjustment of the levelness of the to-be-detected lens is realized; the problem that the levelness of the to-be-detected lens is difficult to guarantee during detection is solved, and the levelness of the to-be-detected lens is adjusted conveniently.

This application provides an imaging method, apparatus, and computer device for a target region. The method includes: acquiring multiple first time-domain data in response to the completion of scanning of the target region by an ultrasonic device; determining first wavenumber domain data corresponding to each first time-domain data; determining a target offset corresponding to each first wavenumber domain data based on the difference between the first wavenumber and the second wavenumber in the first wavenumber domain data; the signal coherence and noise information between the first time-domain signal and the second time-domain signal in the first time-domain data are inversely proportional to the target offset; determining a target weight matching the target offset for each first wavenumber domain data; the target weight is directly proportional to the signal coherence; weighting the first wavenumber domain data based on the target weight to obtain second wavenumber domain data; obtaining an imaging image of the target region based on image domain transformation processing of multiple second wavenumber domain data; and using the imaging image to identify defects in the target region.