Unlock instant, AI-driven research and patent intelligence for your innovation.
Data set making method for deep learning attitude estimation
What is Al technical title?
Al technical title is built by PatSnap Al team. It summarizes the technical point description of the patent document.
An attitude estimation and production method technology, applied in image data processing, calculation, instruments, etc., can solve the problems of complicated production process and difficulty in obtaining 3D models.
Active Publication Date: 2020-02-28
FOSHAN INST OF INTELLIGENT EQUIP TECH
View PDF7 Cites 0 Cited by
Summary
Abstract
Description
Claims
Application Information
AI Technical Summary
This helps you quickly interpret patents by identifying the three key elements:
Problems solved by technology
Method used
Benefits of technology
Problems solved by technology
[0003] The first step in attitude estimation is a high-precision sample data set, but the production process of the traditional LineMod format data set is complicated and cumbersome, especially the precise 3D model required in the data set is difficult to obtain, especially for irregular objects
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more
Image
Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
Click on the blue label to locate the original text in one second.
Reading with bidirectional positioning of images and text.
Smart Image
Examples
Experimental program
Comparison scheme
Effect test
Embodiment 1
[0027] The target object in this embodiment is a milk carton as an example.
[0028] S1: Prepare a camera for collecting photos, use the checkerboard to calibrate the camera, obtain the camera's internal parameter mtx and external parameter dist, and also determine the camera coordinate system, because the camera coordinate system is based on the optical center of the camera as the origin;
[0029] S2: Generate and print a code disc including at least one QR code, and place the milk carton on the plane of the code disc; the QR code is generated by the aruco library in opencv, and the printing effect of the code disc is as follows figure 2 shown;
[0030] S3: Use the camera to collect pictures of milk cartons and code discs, at least one QR code in the pictures is not covered by the milk carton; the effect of the collected pictures is as follows image 3 shown.
[0031] S4: Identify the QR code in the picture, select a QR code that is not covered by the milk carton and set u...
Embodiment 2
[0038] Further, a vertex order detection step is also included. After the pixel coordinates of the 8 vertices are calculated in step S6, the order of the 8 vertices is proposed, and the order of the vertices mapped on the milk carton is checked by reprojection to see if the order is consistent with the proposed order. If not , adjust to be consistent, and then execute step S7. Such as Figure 4 As shown, first make sure that the labels of the 8 vertices are 1-8 respectively, and set a label for each vertex, if Figure 4 The right side is set as the front side, and the milk carton was placed upside down when the picture was collected, then the fixed point 1 at this time corresponds to the original 5, 3 corresponds to 7, 2 corresponds to 6, and 4 corresponds to 8, then the point at this time The correct order should be 56781234, just adjust the order number of vertices from 1-8 to 56781234.
Embodiment 3
[0040] In step S2 of this embodiment, a code plate including two two-dimensional codes is generated and printed, and the milk carton is placed between the two two-dimensional codes. This embodiment is mainly to ensure that at least one two-dimensional code is not covered by the milk carton when the picture is captured, so that it can be captured and identified.
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More
PUM
Login to View More
Abstract
The invention relates to a data set making method, in particular to a data set making method for deep learning attitude estimation. The problems that a three-dimensional model needs to be manufacturedto obtain three-dimensional information of a target in traditional LineMod format data set manufacturing, and manufacturing of a three-dimensional model of an irregular object is very complex and nothigh in precision are solved. The three-dimensional coordinate information of the target object is obtained by using the information recognized by the two-dimensional code and the size of the minimumbounding rectangle of the target object, the manufacturing of a three-dimensional model is avoided. A set of standard data set manufacturing process is planned, and the method has important use significance for the application of deep learning in the aspect of attitude estimation. According to the method, the deep learning model yolo6d is used for training and testing the data set, and the finalaccuracy is higher than 95%.
Description
【Technical field】 [0001] The invention relates to a data set production method, in particular to a data set production method for deep learning attitude estimation. 【Background technique】 [0002] Pose estimation has always been an indispensable research content for researchers in the field of vision. Pose estimation is the basis of many spatial tasks and the premise of mobile robot movement and robot grasping. Traditional pose estimation algorithms have been developed for decades and have achieved Achievements have also encountered many technical bottlenecks. The rise of deep learning has given new ideas to attitude estimation. With the development of deep learning, the recognition ability of the model has become stronger and stronger. The two-dimensional target detection technology has been very mature, and the three-dimensional attitude estimation algorithms have emerged in an endless stream in recent years. It has been perfected day by day and has great practical value....
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More
Application Information
Patent Timeline
Application Date:The date an application was filed.
Publication Date:The date a patent or application was officially published.
First Publication Date:The earliest publication date of a patent with the same application number.
Issue Date:Publication date of the patent grant document.
PCT Entry Date:The Entry date of PCT National Phase.
Estimated Expiry Date:The statutory expiry date of a patent right according to the Patent Law, and it is the longest term of protection that the patent right can achieve without the termination of the patent right due to other reasons(Term extension factor has been taken into account ).
Invalid Date:Actual expiry date is based on effective date or publication date of legal transaction data of invalid patent.
Login to View More 
Login to View More