Refrigerator trundle detection method, system and device based on visual identification
A technology for visual recognition and detection methods, applied in neural learning methods, image data processing, instruments, etc., can solve problems such as wrong installation, caster bolts, casters missing, etc., to reduce wrong installation, improve efficiency, and reduce manpower. The effect of the consumption of capital
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0057] Such as figure 1 As shown, a method for detecting casters of refrigerators based on visual recognition, the operation process is as follows:
[0058] Step 1, collect the pictures of the produced freezer bottom plate, set up the freezer caster data set and the caster bolt data set, and establish the freezer bottom plate database;
[0059] Step 2, use the algorithm to expand and enhance the number of freezer floor pictures in the database, and use manual labeling;
[0060] Step 3, split the marked data into training set and verification set, perform training and verification, generate multiple types of floor inspection models, and store the floor inspection models in the inspection system;
[0061] Step 4, control the industrial camera to take pictures of the bottom plate of the freezer to be tested, and obtain the picture of the bottom plate;
[0062] Step 5, preprocessing the picture of the bottom plate of the freezer obtained in step 4, obtaining the minimum circumsc...
Embodiment 2
[0073] A visual recognition-based detection system for refrigerator casters, including:
[0074] Acquisition module, storage module, expansion module, test module, camera module, preprocessing module, judgment module and display module;
[0075] Wherein the collection module is used to collect the pictures of the produced freezer bottom plate, and send the collected data to the storage module;
[0076] The storage module receives the bottom plate data sent by the acquisition module, organizes the data, and establishes a refrigerator bottom plate database;
[0077] The expansion module calls the bottom plate data in the database, uses an algorithm to expand the number of bottom plate pictures, and uses manual marking, and sends the marked data to the verification module;
[0078] The test module receives the marked expansion data sent by the expansion module, splits it into a training set and a verification set, performs training and verification again, generates a base plate ...
Embodiment 3
[0090] A device for detecting caster wheels of a refrigerator based on visual recognition, comprising: a processor and a memory, the processor is coupled to the memory, a program for detecting caster wheels of a refrigerator based on visual recognition is stored in the memory, and the processor executes the The freezer caster detection program based on visual recognition is to realize any one based on visual recognition caster detection method of freezer as claimed in claims 1-8;
[0091] In the specific implementation process, the pictures of the bottom plates of various types of refrigerators produced by the company are collected, the database of the bottom plates of the refrigerators is established, and the number of pictures of the bottom plates of the refrigerators in the database is expanded and marked by the algorithm, and they are divided into training sets and verification sets for training and verification. Generate different types of floor inspection models, start th...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More