FIELD: computer equipment.
SUBSTANCE: method of detecting and classifying small objects on images obtained by synthetic aperture radar stations involves: detecting and classifying regions with potential objects based on a convolutional neuron network of an architecture such as a detection network (DetectNet), wherein detection of areas with potential objects and their preliminary classification using RI of low resolution, wherein final classification of objects is carried out using a second neuron network of residual network type (ResNet), which uses radar images with high resolution, wherein merging of two neural networks into common system is performed by method of increasing resolution for detected areas of interest by interpolation, and interpolation may be both fixed – bicubic and adaptive – trainee.
EFFECT: high accuracy of classifying objects on a radar image.
1 cl, 1 dwg
| Title | Year | Author | Number |
|---|---|---|---|
| HARDWARE-SOFTWARE COMPLEX DESIGNED FOR PROCESSING AERIAL PHOTOGRAPHS IN VISIBLE AND FAR INFRARED BAND FOR DETECTION, LOCALISATION AND CLASSIFICATION OF BUILDINGS OUTSIDE OF LOCALITIES | 2020 |
|
RU2752246C1 |
| SOFTWARE AND HARDWARE COMPLEX DESIGNED FOR PROCESSING AEROSPACE IMAGE OF TERRAIN FOR PURPOSE OF DETECTION, LOCALIZATION AND CLASSIFICATION BY TYPE OF AVIATION AND LAND EQUIPMENT | 2021 |
|
RU2811357C2 |
| METHOD FOR AUTOMATED ANALYSIS OF DIGITAL FLUOROGRAPHY IMAGES | 2018 |
|
RU2684181C1 |
| OBJECT RECOGNITION METHOD IN VIDEO SURVEILLANCE SYSTEM | 2022 |
|
RU2788301C1 |
| METHOD FOR PROVIDING COMPUTER VISION | 2022 |
|
RU2791587C1 |
| METHOD FOR AUDIOVISUAL RECOGNITION OF PERSONAL PROTECTION EQUIPMENT ON HUMAN FACE | 2022 |
|
RU2791415C1 |
| METHOD OF FORMATION OF NEURAL NETWORK ARCHITECTURE FOR CLASSIFICATION OF OBJECT TAKEN IN CLOUD OF POINTS, METHOD OF ITS APPLICATION FOR TEACHING NEURAL NETWORK AND SEARCHING SEMANTICALLY ALIKE CLOUDS OF POINTS | 2017 |
|
RU2674326C2 |
| METHOD FOR PROCESSING IMAGES BY CONVOLUTIONAL NEURAL NETWORKS | 2020 |
|
RU2771442C1 |
| METHOD FOR CREATING COMBINED NEURAL NETWORK CASCADES WITH COMMON FEATURE EXTRACTION LAYERS AND WITH MULTIPLE OUTPUTS, TRAINED ON DIFFERENT DATASETS SIMULTANEOUSLY | 2021 |
|
RU2779408C1 |
| SYSTEM AND METHOD FOR ARTIFICIAL NEURAL NETWORK INVARIANT TO TRANSFERRING | 2017 |
|
RU2656990C1 |
