Method and apparatus for obtaining seed tissue samples based on vision module laser cutting
By combining visual modules and laser cutting technology with deep learning to identify seed features, precise cutting of seed tissue samples was achieved, solving the problems of seed damage and inaccurate sampling in traditional methods, and improving the success rate of seed cutting and the seed viability retention rate.
CN122149904APending Publication Date: 2026-06-05WUWEN (GUANGZHOU) TECHNOLOGY CO LTD
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- WUWEN (GUANGZHOU) TECHNOLOGY CO LTD
- Filing Date
- 2025-12-05
- Publication Date
- 2026-06-05
AI Technical Summary
Technical Problem
Traditional methods are prone to damaging seeds during seed cutting and sampling, and are difficult to adapt to different seed shapes and endosperm locations, resulting in inaccurate sampling and low success rates.
Method used
A vision-based laser cutting method is adopted, which uses seed images for path planning and epidermal removal planning. By using non-contact laser cutting technology and deep learning to identify seed features, the accuracy and success rate of cutting are ensured.
Benefits of technology
It achieves precise cutting of seeds, maximizing the preservation of seed vigor and germination potential, improving sampling success rate and accuracy, and avoiding mechanical stress damage.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN122149904A_ABST
Abstract
The application discloses a method and device for obtaining seed tissue samples based on a visual module, and comprises the following steps: obtaining an original image of a seed in a fixing device; pre-processing the original image; extracting a target from the processed image; planning a cutting path or an epidermis removal area for the extracted target; and cutting the seed fixed in the clamping jig by a cutting assembly to obtain a tissue sample. After the seed is fixed by the clamping jig, the visual module is used to obtain a seed image, phenotype recognition is performed on the seed, the embryo and endosperm of the seed are recognized, then the seed is planned with a cutting path and an epidermis removal area according to the cutting target, and the laser is guided to complete cutting or surface treatment, so that the cutting path can be generated in real time for each seed with different shapes and sizes.
Need to check novelty before this filing date? Find Prior Art