Method for converting optical image into strip mode SAR radar image, electronic equipment and storage medium

Abstract

The invention discloses a method for converting an optical image into a stripe mode SAR radar image, electronic equipment and a storage medium. The method comprises the following steps: acquiring an optical image frame in visual software and preprocessing the optical image frame; establishing an inter-frame position relationship through image registration; converting the image to a frequency domain, and performing low-pass filtering and noise addition to simulate SAR image characteristics; post-processing the image and marking a total reflection area; and finally, segmenting the image according to a strip mode and sequentially displaying the image. Based on the traditional image processing and feature registration technology, efficient and stable simulation of SAR image features is achieved, and the method has the advantages of being high in response speed, high in anti-interference capacity, high in image authenticity and the like and is suitable for unmanned aerial vehicle simulation training and radar image simulation systems.

Description

A method, electronic device, and storage medium for converting optical images to striped SAR radar images. Technical Field

[0001] This invention belongs to the field of image processing technology, and in particular relates to a method, electronic device and storage medium for converting optical images to strip mode SAR radar images. Background Technology

[0002] In UAV simulation training systems, optical images are typically used to construct visual scenes to provide realistic visual effects. However, in actual combat, UAVs often carry multiple sensors, such as SAR radar, whose imaging mechanisms differ significantly from optical images. Currently, while AI-based image style transfer methods can generate radar images with good visual effects, they suffer from problems such as strong dependence on training data, uncontrollable generation results, and poor real-time performance. Traditional image processing methods, although stable and efficient, lack accurate simulation of the unique textures, noise, and stripe imaging patterns of SAR images. Therefore, there is an urgent need for an optical image conversion method that can guarantee real-time performance while accurately simulating the characteristics of SAR images. Summary of the Invention

[0003] To address the problems existing in the prior art, the present invention provides the following technical solution: a method for converting optical images to striped SAR radar images, comprising: communicating with visual software and acquiring video stream image frames from an aircraft camera as image frames to be processed; performing preprocessing operations on the image frames; constructing a temporal relationship model between image frames; predicting the relative positional relationship between optical images based on the preprocessed image frames using an image registration calculation method; converting the preprocessed image frames into frequency domain images, filtering out high-frequency information and adding texture noise using a low-pass filter, and then converting them into spatial domain images; and performing post-processing on the images according to actual rule requirements, marking total reflection areas, and converting them into striped images.

[0004] Furthermore, the preprocessing operation includes at least one of the following operations: normalizing the shape and size of the image; aligning or anchoring the image position; and converting the color image to a grayscale image.

[0005] Furthermore, the image registration calculation method includes: extracting feature points using the Harris corner detection algorithm; extracting and matching features using the Scale Invariant Feature Transform (SIFT) algorithm; or employing a feature matching method based on deep learning.

[0006] Furthermore, the low-pass filter is a circular ideal low-pass filter, and its frequency domain transfer function is defined as: in, For frequency domain coordinates, This is the cutoff frequency radius, with a default value of 30 pixels.

[0007] Furthermore, in the method for calculating texture noise, the noise addition process occurs in the frequency domain, and the noise matrix is... ,in, and These represent the width and height of the image matrix, and each element... ,among A random value between 0 and 255. This represents the noise intensity, with a default value of 0.1.

[0008] Furthermore, post-processing operations performed on the image include reducing overall brightness and averaging the data of each pixel channel.

[0009] Furthermore, the method for marking total reflection regions includes: confirming total reflection region data, calculating the matrix to identify image frames, and ensuring that each identified position has a minimum interval.

[0010] The present invention also provides an electronic device, the electronic device comprising: a memory storing executable instructions; and a processor executing the executable instructions in the memory to implement a method for converting an optical image to a striped SAR radar image as described above.

[0011] The present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements a method for converting an optical image to a striped SAR radar image as described above.

[0012] Beneficial effects: Based on traditional image processing and feature registration techniques, this invention achieves efficient and stable simulation of SAR image features, with advantages such as fast response speed, strong anti-interference ability, and high image realism. It is suitable for UAV simulation training and radar image simulation systems. Attached Figure Description

[0013] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention. For those skilled in the art, other related drawings can be obtained from these drawings without creative effort.

[0014] Figure 1 shows a flowchart illustrating the steps of converting an optical image into a striped SAR radar image according to an embodiment of the present invention.

[0015] Figure 2 shows a flowchart of the steps for displaying a striped SAR image interface according to an embodiment of the present invention.

[0016] Figure 3 shows an image deformation effect diagram of an embodiment of the present invention.

[0017] Figure 4 shows a rendering of an image stripe display according to an embodiment of the present invention. Detailed Implementation

[0018] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0019] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0020] Figure 1 shows a flowchart illustrating the steps of converting an optical image into a striped SAR radar image according to an embodiment of the present invention.

[0021] As shown in Figure 1, the method for converting the optical image into a strip-mode SAR radar image includes: Step 101, acquiring the image frames to be processed in the visual software and recording them in the order of reception time; Step 102, using corner detection for registration, constructing the relative position changes of multiple images, and recording the relative positions; Step 103, performing preprocessing operations on the received optical image, unifying the shape and size, and converting it into a grayscale image; Step 104, using Fourier transform to convert the image from a spatial domain image to a frequency domain image, simulating the SAR radar image information collection process; Step 105, using a low-pass filter to filter out high-frequency information in the frequency domain image and adding noise to simulate the texture of the radar image; Step 106, using inverse Fourier transform to convert the frequency domain image into a spatial domain image; Step 107, performing post-processing on the spatial domain image to reduce the overall brightness, average the data of each pixel channel, and blur the gaps between images; Step 108, marking the corner reflector structure areas in the image to create a starlight-like flickering effect; Step 109, segmenting the image according to the relative position and displaying the image as a strip structure.

[0022] In one example, in the visual software, the aircraft flies in a straight line, acquires and transmits ground optical image information; in another example, the registration method can be Harris corner detection or Scale Invariant Feature Transform (SIFT), depending on the real-time requirements.

[0023] Figure 2 shows a flowchart of a strip-patterned SAR radar image displayed line by line according to an embodiment of the present invention.

[0024] As shown in Figure 2, the steps for displaying the strip pattern SAR radar image include: Step 109, segmenting the converted image to obtain a strip structure image with consistent length and width.

[0025] Step 110: Read the information of the next image, and proceed to different steps depending on whether the image information is read; Step 111: If the information of the next image is not read, wait for the image to be received, and then proceed to Step 101; Step 112: If the information of the next image is read, accumulate the relative positions of the current strip image (to the first image) based on the relative positions of the neighboring images calculated in Step 109.

[0026] Step 113: Based on the display requirements of the strip mode, display the next strip image at regular intervals according to the time, and adjust the position of all images on the screen.

[0027] Specifically, after the radar is activated, it acquires each frame of image transmitted by the visual software, with the image position derived from the interaction information between the radar and the aircraft. Ground image data is collected in steps of 300 pixels, depending on the drone's flight altitude. Each frame is converted into a radar image through a low-pass filter. Multiple images in the scene are stitched together to obtain the displacement difference between adjacent images. Imaging stops during the drone's turning and hovering process.

[0028] In the image stitching process, choosing a suitable method is crucial. Practice has shown that most commonly used stitching methods, such as convolutional neural networks and a series of traditional methods, can achieve the desired algorithmic results. Considering time efficiency and the simulated scenario, the SIFT method from the traditional approach is currently adopted. Regardless of the method, distortion will inevitably occur during image stitching; the more images stitched, the greater the distortion. Completely eliminating distortion requires more images and a longer stitching time. Therefore, slight distortion is accepted. Since radar images are grayscale, the overall brightness is reduced, and the gaps between the striped images are blurred. Figure 3 shows the effect of distortion on image stitching.

[0029] Frequency domain image is ,in For frequency domain images, For Fourier transform, To represent the Fourier transform operator, This is a spatial domain image. Low-pass filter. For a circular filter, the calculation method is as follows: radius The default value is 30 pixels, which determines the amount of low-frequency components to retain. The filter is a circular filter.

[0030] filter Compared with the transformed image Multiplication yields the filtered image. : Finally, the filtered spatial domain image is obtained through inverse Fourier transform. : in, This is the inverse Fourier transform.

[0031] In the method for calculating texture noise, the noise addition process occurs in the frequency domain, and the noise is calculated as follows: ,in, , For the width and height of the image matrix, each element ,among A random value between 0 and 255. This represents the noise intensity, with a default value of 0.1.

[0032] Figure 4 shows the final imaging result of a scene image according to an embodiment of the present invention. In the image, a starlight-like twinkling effect is obtained through calculations between matrices. After selecting a determined position of total internal reflection angle, on a 10×10 matrix, the pixel values ​​of the diagonals are increased by 200, and those exceeding 255 are calculated as 255.

[0033] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for converting optical images to strip-mode SAR radar images, characterized in that, include: The system communicates with the visual software and acquires video stream image frames from the aircraft camera as image frames to be processed, and performs preprocessing operations on the image frames. Construct a temporal relationship model between image frames; Based on the preprocessed image frames, the relative positional relationship between optical images is predicted using image registration calculation methods; the preprocessed image frames are converted into frequency domain images, and high-frequency information is filtered out by a low-pass filter, texture noise is added, and then converted into spatial domain images. Based on the actual rules, the image is post-processed, the total reflection area is marked, and the image is converted into a striped pattern.

2. The method for converting optical images to strip-mode SAR radar images as described in claim 1, characterized in that, The preprocessing operation includes at least one of the following operations: normalizing the shape and size of the image; aligning or anchoring the image position; and converting the color image to a grayscale image.

3. The method for converting optical images to strip-mode SAR radar images as described in claim 2, characterized in that, The image registration calculation method includes: extracting feature points using the Harris corner detection algorithm; extracting and matching features using the Scale Invariant Feature Transform (SIFT) algorithm; or using a feature matching method based on deep learning.

4. The method for converting an optical image to a strip-mode SAR radar image as described in claim 1, characterized in that, The low-pass filter is a circular ideal low-pass filter, and its frequency domain transfer function is defined as: in, For frequency domain coordinates, This is the cutoff frequency radius, with a default value of 30 pixels.

5. The method for converting an optical image to a strip-mode SAR radar image as described in claim 1, characterized in that, In the method for calculating texture noise, the noise addition process occurs in the frequency domain, and the noise matrix is... ,in, and These represent the width and height of the image matrix, and each element... ,among A random value between 0 and 255. This represents the noise intensity, with a default value of 0.

1.

6. The method for converting an optical image to a strip-mode SAR radar image as described in claim 1, characterized in that, Post-processing operations performed on the image include reducing overall brightness and averaging the data of each pixel channel.

7. The method for converting an optical image to a strip-mode SAR radar image as described in claim 1, characterized in that, The method for marking total reflection regions includes: confirming the total reflection region data, calculating the matrix to identify the image frame, and ensuring that each marked position has a minimum interval.

8. An electronic device, characterized in that, The electronic device includes: a memory storing executable instructions; and a processor that executes the executable instructions in the memory to implement a method for converting an optical image to a striped SAR radar image according to any one of claims 1-7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements a method for converting an optical image to a striped SAR radar image according to any one of claims 1-7.

Images