Image Super-Resolution

Capability Introduction

9x magnification, doubled clarity

Based on the wide application of deep learning in the computer vision domain, an image can be intelligently zoomed in, or the compressed noise can be removed without changing the resolution, so as to obtain a clearer, sharper, and cleaner image than that obtained from conventional image processing. This algorithm is based on the deep neural network, and relies on Huawei NPU chips. Compared with pure CPU computing, it improves computing speed by nearly 50 times. This API is built into Huawei smartphones. It consumes less ROM and RAM than alternatives, effectively reducing the size of apps, saving memory in the process.

Less traffic, better image quality

This API intelligently zooms in on images with the goal of enhancing image quality and improving the viewing experience, as well as improving image compression and transmission.

Image compression and quality enhancement

To save storage space and traffic via high image compression. To enhance image quality and transmission speed via 1x super-resolution.

Improved zoom-in images

3x super-resolution furnishes images with clearer effects, reducing JPEG compression noises, and thus improving image browsing experience.


ImageSuperResolution superResolution = new ImageSuperResolution(mContext);
Frame frame = new Frame();
ImageResult result = superResolution.doSuperResolution(frame, null);
Bitmap resultBitmap = result.getBitmap();


The image after super resolution

Start development here with Huawei

Service introduction

Service overview, application scenarios, and functions
Learn more

Access preparation

Preparation for access, such as registration, authentication, and agreement signing
Learn more

Sample code

Demonstration of sample code
Learn more

API description

Interface parameter definition, description, restrictions, and constraints
Learn more
1Are extra management resources needed when using super-resolution API?

Huawei super-resolution API encapsulates all additional resource management into the interface. Users do not need to manage any additional resources, and can rely on existing resources.

2How can I troubleshoot for an abnormally running super-resolution API?

1. Refer to the Android error logs. This is the fastest and most direct way to find the source of the error. 2. Refer to the result code to determine the general error type. 3. In most cases, the main cause of a running error is an invalid input. Read the restrictions and constraints in the API document and check carefully whether the input parameters meet the requirements. 4. Check whether the app has the corresponding permission, for example, storage permission.

3Why is the Huawei super-resolution function still available in the interface, when it does not support quality selection?

1. Super-distribution quality selection for the Huawei super-resolution API is planned to be made available in the future. Quality selection provides an optimal solution for users. A higher quality super-resolution means a larger computing workload, which results in higher power consumption and longer latency. 2. In some scenarios, the quality of the super-resolution is not very sensitive, or varies according to the user. For example, the difference between the highest quality and a substantially lower quality, is only noticeable to experts in the field, or when the image is magnified by 100%. However, the running time can vary greatly.