FIELD: building demolition equipment.
SUBSTANCE: group of inventions relates to the field of building demolition equipment and can be used for the material segmentation for building destruction scenarios. Method comprises steps of: inputting the image of the target building destruction scenario to be segmented into the trained building destruction scenario material segmentation model to obtain the material segmentation result; wherein obtaining the building destruction scenario material segmentation model includes the stages, at which: obtaining the road scenario material data set; converting the segmentation images of the road scene material into a set of data on the road scene material to obtain a set of data on the building scene material; adding material objects of destruction area and corresponding markers of material on image of set of data on building scenario material to obtain a set of data on building destruction scenario material; second deep semantic segmentation model is trained using a set of data on the building destruction scenario material to obtain a basic segmentation model of the building destruction scenario material; and performing domain adaptive training on the basic segmentation model of the material of the building destruction scenario using the marked images in the images of the sequence of the target scenario to obtain a trained segmentation model of the material of the building destruction scenario.
EFFECT: high accuracy of segmenting material of a building being demolished.
11 cl, 5 dwg
Title | Year | Author | Number |
---|---|---|---|
SYSTEM FOR IDENTIFYING FRESH PRODUCTS FOR A RETAIL TRADE CASH TERMINAL | 2018 |
|
RU2769888C2 |
TRAINING GAN (GENERATIVE ADVERSARIAL NETWORKS) TO CREATE PIXEL-BY-PIXEL ANNOTATION | 2019 |
|
RU2735148C1 |
METHOD AND DEVICE FOR CONTROLLING RISKS FOR ROBOTS FOR DEMOLITION OF STRUCTURES | 2024 |
|
RU2837975C1 |
VISUALIZATION OF RECONSTRUCTION OF 3D SCENE USING SEMANTIC REGULARIZATION OF NORMALS TSDF WHEN TRAINING NEURAL NETWORK | 2023 |
|
RU2825722C1 |
METHOD AND DEVICE FOR TRAINING A FACE RECOGNITION MODEL AND A DEVICE FOR DETERMINING THE KEY POINT OF THE FACE | 2019 |
|
RU2770752C1 |
INTERACTIVE IMAGE SEGMENTATION | 2023 |
|
RU2833268C1 |
IMAGE PROCESSING DEVICE AND METHOD AND EQUIPMENT | 2019 |
|
RU2794062C2 |
METHOD FOR ESTIMATING THE DEPTH OF A SCENE BASED ON AN IMAGE AND COMPUTING APPARATUS FOR IMPLEMENTATION THEREOF | 2020 |
|
RU2761768C1 |
METHOD, EQUIPMENT AND DEVICE FOR IMAGE PROCESSING | 2019 |
|
RU2791810C2 |
RAPID TWO-LAYER NEURAL NETWORK SYNTHESIS OF REALISTIC IMAGES OF A NEURAL AVATAR BASED ON A SINGLE IMAGE | 2020 |
|
RU2764144C1 |
Authors
Dates
2025-04-28—Published
2024-09-26—Filed