18455918. IMAGE PROCESSING METHOD AND RELATED APPARATUS simplified abstract (HUAWEI TECHNOLOGIES CO., LTD.)
Contents
IMAGE PROCESSING METHOD AND RELATED APPARATUS
Organization Name
Inventor(s)
IMAGE PROCESSING METHOD AND RELATED APPARATUS - A simplified explanation of the abstract
This abstract first appeared for US patent application 18455918 titled 'IMAGE PROCESSING METHOD AND RELATED APPARATUS
Simplified Explanation
The patent application describes an image processing method for artificial intelligence applications. Here is a simplified explanation of the abstract:
- The method involves processing an input image by obtaining an input feature map, which consists of two sub-feature maps with different resolutions.
- A target network is used to perform feature fusion processing on the input feature map, combining features from the higher-resolution sub-feature map with the lower-resolution sub-feature map.
- The fusion process occurs from a low level to a high level in the target network, enhancing the features of the image.
- Object detection is then performed on the processed image using the output feature map, resulting in an object detection result.
Potential applications of this technology:
- Object detection and recognition in various fields such as autonomous driving, surveillance systems, and robotics.
- Image analysis and understanding in medical imaging, where accurate detection of objects or anomalies is crucial.
- Video analysis and tracking, enabling real-time object detection and tracking in video streams.
Problems solved by this technology:
- Enhances the accuracy and efficiency of object detection by fusing features from different resolution sub-feature maps.
- Improves the ability to detect objects at different scales and levels of detail.
- Enables better understanding and interpretation of images by combining features from different levels of abstraction.
Benefits of this technology:
- Improved object detection accuracy and reliability.
- Enhanced performance in scenarios with objects at different scales or resolutions.
- Increased efficiency and speed in image processing tasks.
- Enables more advanced and accurate AI applications in various fields.
Original Abstract Submitted
An image processing method is disclosed in embodiments of this disclosure and is applied to the field of artificial intelligence. The method includes: obtaining an input feature map of an image to be processed, where the input feature map includes a first input sub-feature map and a second input sub-feature map, and resolution of the first input sub-feature map is higher than resolution of the second input sub-feature map; performing feature fusion processing on the input feature map by using a target network, to obtain an output feature map, where a feature of the first input sub-feature map is fused to a feature of the second input sub-feature map from a low level to a high level in the target network; and performing, based on the output feature map, object detection on the image to be processed, to obtain an object detection result.
