A main fruit stem depth-based stacked string type fruit grabbing priority determination method for a parallel robot
A determination method and robot technology, applied in the fields of instruments, computer parts, character and pattern recognition, etc., can solve the problems of difficult grasping and detection, irregular distribution of fruit stems and fruit grains, etc., to improve learning ability, stable and reliable grasping. take effect
Active Publication Date: 2019-05-31
JIANGSU UNIV
View PDF3 Cites 2 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Compared with independent fruits such as apples, pears, pineapples, etc., stacked string fruits such as grapes, longan, litchi, etc. are due to the irregular distribution of fruit stems and fruit grains, the lack of shape and position constraints of the main fruit stem, and the variety of fruit cluster shapes. , its grasp detection based on machine vision is still difficult
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment
[0094]The present invention emphatically proposes a method for determining the priority of stacked string fruit grasping based on the depth of the main fruit stem for parallel robots, which solves the problem that all depth values of each main fruit stem can form a depth set, which makes it difficult to judge The depth of the main fruit stem and further determine the priority of grasping based on the depth. A depth reference object was designed to construct a pre-training data set with uniform depth distribution, and a multi-transfer learning training strategy was designed to improve the network's ability to learn the depth characteristics of the main fruit stems of stacked string fruits. After that, the convolutional layer with no scale change, the pooling layer with smaller scale change and multiple fully connected layers are designed, and a classification model of the main fruit stem depth set based on the convolutional neural network architecture with less pooling and mul...
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 discloses a main fruit stem depth-based stacked string type fruit grabbing priority determination method for a parallel robot. The method includes: under a parallel robot fruit sorting system, constructing a stereoscopic vision detection system for stacking bunched fruits based on a Kinect sensor; acquiring three-dimensional visual information of stacked bunchy fruits, designing a depth reference object to construct a pre-training data set, constructing a main fruit stem depth data set of the stacked bunchy fruits, expanding the data set, and increasing the distribution range ofthe data set; constructing a main fruit stem depth set grading model, and increasing the characteristic quantity of the main fruit stem depth; And designing a multi-migration learning training strategy to train the network, performing visual analysis and precision testing on the network, adjusting parameters, and performing training for multiple times until the precision meets the requirement, thereby realizing deep set grading of the stacked bunchy fruits. Finally, the grabbing priority of the stacked bunchy fruits is accurately determined, and a foundation is laid for the parallel robot to accurately, quickly and nondestructively automatically sort the stacked bunchy fruits.
Description
technical field [0001] The invention relates to the field of machine vision, in particular to a method for determining the grasping priority of stacked string fruits based on machine vision, image processing and convolutional neural network. Background technique [0002] In recent years, my country's fruit production has grown rapidly, and traditional manual sorting methods have been difficult to meet the needs of modern agricultural production. Automatic fruit sorting based on robotic technology is of great importance to the automation, scale, and precision development of agricultural production and agricultural product processing. significance. In the robot-based automatic fruit sorting process, the accurate grasping detection of fruits is the prerequisite for the robot to achieve accurate, fast and non-destructive grasping control. Due to the advantages of non-contact, strong applicability, and high cost performance, machine vision is suitable for solving the problem of g...
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
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/00G06K9/32G06K9/62
Inventor 高国琴张千
Owner JIANGSU UNIV