Unlock instant, AI-driven research and patent intelligence for your innovation.

Dish image background stripping method

A dish and image technology, applied in the field of smart catering, can solve problems such as incompleteness and inaccurate data, and achieve the effect of enhancing recognition ability and solving inaccurate data

Pending Publication Date: 2020-05-15
SHANDONG SYNTHESIS ELECTRONICS TECH
View PDF14 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The technical problem to be solved by the present invention is to provide a method for stripping the background of dish images, which solves the problem of inaccurate and incomplete data generated in the stage of dish recognition data acquisition, and enhances dish recognition learning network recognition by generating a single non-vessel dish data information ability

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
  • Dish image background stripping method
  • Dish image background stripping method
  • Dish image background stripping method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0021] This embodiment discloses a method for stripping the background of a dish image, and the method includes the following steps:

[0022] S01), constructing a dish positioning sample database, the dish positioning sample database contains images of utensils with multiple sources and containing dishes;

[0023] In this embodiment, the image sources of the dish location sample database include photos taken by cameras, cameras or images intercepted from videos taken by them, and images obtained from the network.

[0024] In order to increase sample diversity, in this embodiment, the data in the positioning sample database is cleaned, enhanced, rotated, mirrored, distorted, and blurred.

[0025] In order to facilitate the training of the deep learning target positioning model in the next step, in this embodiment, the data in the sample database is marked, and the marked content is the circumscribed rectangle of the space where the vegetable serving vessel is located.

[0026]...

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

No PUM Login to View More

Abstract

The invention discloses a dish image background stripping method. The method comprises the steps: generating a dish-free single vessel image by adopting a deep learning and edge detection combined method, and acquiring a pure dish region image by subtracting the dish-free single vessel image from a single dish image, so the invalid background can be removed, the content of the core target region can be extracted, and further training of machine learning is facilitated. Therefore, the problem that the data generated in the dish identification data acquisition stage is inaccurate and incompletecan be solved, and the dish identification learning network identification capability is enhanced by generating single vessel-free dish data information.

Description

technical field [0001] The invention relates to the field of smart catering, in particular to a method for stripping the background of a dish image. Background technique [0002] In the process of dish image recognition, the target recognition is usually performed after the target is extracted from the whole dish container, and there is noise influence of the container on the target recognition in this process. At the same time, the target positioning error caused by the problem of the utensil itself will also affect the subsequent recognition work. Therefore, the detection and extraction of the dish area is an important technical component of the recognition process. [0003] The patent with the publication number CN201910315763 discloses a dish identification method. In the process of sample learning, only the target dish is roughly positioned to obtain the dish utensil information, and then feature extraction and comparison are performed to determine which dish the curren...

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
no application Login to View More
IPC IPC(8): G06T7/13G06T7/194G06T7/70
CPCG06T7/13G06T7/194G06T7/70G06T2207/10004G06T2207/20081G06T2207/20084
Inventor 瞿晨非方亮许野平王阔阔凌桂婷刘永悦
Owner SHANDONG SYNTHESIS ELECTRONICS TECH