IMAGE STITCHING METHOD, APPARATUS AND DEVICE BASED ON REINFORCEMENT LEARNING AND STORAGE MEDIUM

    公开(公告)号:US20240378839A1

    公开(公告)日:2024-11-14

    申请号:US18767924

    申请日:2024-07-09

    Abstract: The present application provides an image stitching method, apparatus and device based on reinforcement learning and a storage medium. The method includes: acquiring initial calibration parameters, collecting a sample image and position information of a motion platform; setting a negative reward function; acquiring a state set and a negative reward value set according to a randomly generated action set, the initial calibration parameters, the position information of the motion platform and the negative reward function to construct a probability kinematics model; constructing a state value function based on an occurrence probability of the state, and acquiring an optimal action by optimizing the state value function; and acquiring optimized calibration parameters through the optimal action and the initial calibration parameters, and carrying out image stitching on corresponding sample images through the optimized calibration parameters. The application solves the technical problem of low image stitching quality in the prior art.

    THREE-DIMENSIONAL MEASUREMENT METHOD AND RELATED APPARATUS

    公开(公告)号:US20230401730A1

    公开(公告)日:2023-12-14

    申请号:US18154557

    申请日:2023-01-13

    CPC classification number: G06T7/521 G01C3/02 G06T2207/10028

    Abstract: A three-dimensional measurement method comprises: converting a total number of levels of sawtooth fringes into a Gray code and acquiring a sawtooth slope coefficient; fusing the coefficient into a sawtooth fringe image to generate a target sawtooth fringe pattern; projecting each target sawtooth fringe pattern to a surface of a to-be-measured object through a projector, and collecting a deformed target sawtooth fringe pattern on the surface through a camera; solving a Gray code of each sawtooth fringe collection pattern at each pixel point according to a differential property of adjacent pixels in each sawtooth fringe collection pattern and solving a fringe level and a wrapped phase at each pixel point; calculating an absolute phase at each pixel point according to the fringe level and the wrapped phase at each pixel point, and reconstructing a three-dimensional point cloud through triangulation ranging to obtain a three-dimensional model of the to-be-measured object.

    PLANE CODING TARGET AND IMAGE SPLICING SYSTEM AND METHOD APPLYING THE SAME

    公开(公告)号:US20230217036A1

    公开(公告)日:2023-07-06

    申请号:US18065530

    申请日:2022-12-13

    CPC classification number: H04N19/513 H04N19/172

    Abstract: Disclosed are a plane coding target and an image splicing system and method applying the same. The plane coding target comprises a plurality of coding units distributed in an array, the coding unit comprises one central coding point, a plurality of normal coding points and at least one positioning point, and a positioning point distribution style of the positioning point is used for determining coordinates of the central coding point and the normal coding points in a coding unit coordinate system; and coding numerical value sequences of the coding units are different from each other and unique. The plane coding target can realize large-area coding and positioning functions, and the image splicing system applying the plane coding target can solve the problems of splicing error and error accumulation caused by an identification error of a splicing location, thus realizing large-range, high-precision and short-time two-dimensional image splicing.

    TIME OPTIMAL SPEED PLANNING METHOD AND SYSTEM BASED ON CONSTRAINT CLASSIFICATION

    公开(公告)号:US20230185262A1

    公开(公告)日:2023-06-15

    申请号:US17991936

    申请日:2022-11-22

    CPC classification number: G05B19/041 G06F17/16

    Abstract: A time optimal speed planning method and system based on constraint classification. The method comprises: reading path information and carrying out curve fitting to obtain a path curve; sampling the path curve, and considering static constraint to obtain a static upper bound value of a speed curve; considering dynamic constraint, and combining the static upper bound value of the speed curve to construct a time optimal speed model; carrying out convex transformation on the time optimal speed model to obtain a convex model; and solving the convex model based on a quadratic sequence planning method to obtain a final speed curve. The system comprises: a path curve module, a static constraint module, a dynamic constraint module, a model transformation module and a solving module.

    FAST CLUSTERING ALGORITHM BASED ON KERNEL FUZZY C-MEANS INTEGRATED WITH SPATIAL CONSTRAINTS

    公开(公告)号:US20210081827A1

    公开(公告)日:2021-03-18

    申请号:US17088577

    申请日:2020-11-03

    Abstract: A fast clustering algorithm of kernel fuzzy C-means integrated with spatial constraints, including (1) applying the illumination processing algorithm, the preprocessed image affected by illumination is constructed; (2) After step (1), the original image and preprocessed image are mapped to the feature space using Gaussian kernel to cluster and segment. Providing a defect segmentation method for fluorescent glue which is robust to illumination to process and calculate the illuminated image, so as to complete the detection of foreign matters, bubbles and discoloration defects of fluorescent glue in lighting products. The disclosure provides a fast clustering algorithm of kernel fuzzy C-means integrated with spatial constraints. The image is mapped into the feature space, and the objective function of kernel fuzzy C-means clustering is optimized by using the spatial relationship of pixels, so that the clustering process has segmentation robustness to the gray value change of similar pixels caused by environmental changes.

Patent Agency Ranking