Image transmission method and system suitable for geostationary orbit satellite

Abstract

The invention discloses an image transmission method and system suitable for a geostationary orbit satellite. The method mainly comprises the following steps that (1) a sending end collects an image; (2) determining a target edge coordinate; (3) intelligent cutting; (4) encoding the cut image; (5) dynamic fragmentation processing; (6) sending the fragmented data to a receiving end; (7) the receiving end receives the fragmented data and recombines the fragmented data; and (8) the transmitting end retransmits the lost fragment data according to the lost fragment list. According to the technical scheme provided by the invention, through intelligent cutting, dynamic fragmentation and UDP transmission, the characteristics of high delay and limited bandwidth of satellites are adapted, redundant data transmission is reduced, the image transmission efficiency and reliability are improved, and the problems of low transmission efficiency and weak interference resistance of a traditional method are solved.

Description

Technical Field

[0001] This invention relates to the field of satellite communication and image transmission technology, and in particular to an image transmission method and system suitable for geostationary orbit satellites. Background Technology

[0002] Geostationary orbit satellites, due to their fixed orbital altitude, can achieve continuous coverage of specific areas, playing a crucial role in remote sensing monitoring, emergency communications, and meteorological observation. However, their communication links suffer from certain technical bottlenecks: First, high transmission latency (approximately 0.23 seconds per trip), with traditional TCP-based retransmission mechanisms prone to link congestion due to waiting for acknowledgments; second, limited channel bandwidth and poor stability, affected by ionospheric scintillation, rain attenuation, and other factors, resulting in significant fluctuations in link quality and a high likelihood of data packet loss; third, large image data volumes, with traditional image transmission protocols (such as JPEG+HTTP) not optimized for satellite channel characteristics, leading to redundant data transmission further encroaching on limited bandwidth, resulting in low transmission efficiency, image distortion, or excessively high latency.

[0003] In existing technologies, some solutions reduce data volume by simplifying image compression algorithms, but this sacrifices image details; other solutions employ fixed fragmentation transmission strategies, which cannot adapt to dynamic changes in satellite channels, and are prone to overall transmission failure when packet loss rates are high. Therefore, considering the high latency, bandwidth limitations, and unstable channels of geostationary orbit satellites, there is an urgent need for an image transmission method and system that can intelligently reduce invalid data, efficiently compress images, and dynamically adapt to link quality, in order to improve the efficiency and reliability of image transmission. Summary of the Invention

[0004] In view of the shortcomings of the prior art, the purpose of this invention is to provide an image transmission method and system suitable for geostationary orbit satellites, addressing the technical problems of low image transmission efficiency and weak anti-interference capability of geostationary orbit satellites.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: An image transmission method suitable for geostationary orbit satellites includes the following steps: (1) The transmitting end acquires images; (2) Detect the target object in the acquired image and determine the edge coordinates of the target object; (3) Intelligent cropping of the image based on the edge coordinates of the target object, cropping out the invalid area outside the edge coordinates, and preserving the complete shape of the target object; (4) The cropped image is encoded using the HEIF encoding standard; (5) Dynamically adjust the data fragment size according to the current link quality parameters, divide the image data into multiple fragments, and add identification information to each fragment; (6) The sending end transmits the fragmented data to the receiving end via geostationary orbit satellite using the UDP protocol; (7) The receiving end receives the fragmented data, reassembles the fragments according to the identification information, and generates complete image data; if the receiving end detects lost fragmented data, it sends a list of lost fragments back to the sending end. (8) The sending end retransmits the lost fragment data according to the lost fragment list.

[0006] Furthermore, in step (2), the YOLO algorithm is used for target detection.

[0007] Furthermore, in step (2), if there are multiple target objects in the image, each target object is cropped separately without merging or splicing.

[0008] Furthermore, in step (3), the edge-free buffer area is intelligently cropped; after cropping, if the size of the effective area is too large, its resolution is reduced proportionally.

[0009] Furthermore, in step (3), the necessary metadata of the image is retained after intelligent cropping, including camera model, shooting time, and shooting location.

[0010] Furthermore, in step (4), HEIF encoding is controlled by quantization parameters, which are set to 30~40.

[0011] Furthermore, in step (5), when the link quality degrades, the fragment size is reduced to 256 bytes; when the link quality recovers, the fragment size is restored to 512 bytes.

[0012] Furthermore, in step (5), the identification information includes the fragment sequence number, the total number of fragments, the check bit, and the timestamp.

[0013] Furthermore, in step (8), when the sending end retransmits a lost fragment, it prioritizes retransmitting the key fragment, which refers to a fragment containing important information about the target object.

[0014] Accordingly, the present invention provides an image transmission system suitable for geostationary orbit satellites, including a transmitter and a receiver. Unlike the prior art, the transmitter is a UDP client capable of connecting to geostationary orbit satellites and is used to execute steps (1) to (6) and (8) of the method described above; the receiver is a UDP server connected to the terrestrial Internet and is used to execute step (7) of the method described above.

[0015] Beneficial effects: This invention provides an image transmission method and system suitable for geostationary orbit satellites, which has at least the following advantages compared to existing technologies: (1) By intelligently cropping invalid areas and eliminating redundant data, the invalid transmission load is reduced under the limited bandwidth of the satellite, allowing the core content of the image (i.e., the complete shape of the target object) to occupy the transmission resources first; combined with the HEIF encoding standard, the image is efficiently compressed while retaining key information, adapting to the limited bandwidth characteristics of the satellite, thereby solving the technical problems of low transmission efficiency and weak anti-interference ability of geostationary orbit satellite images.

[0016] (2) Furthermore, the low latency advantage of UDP is used to adapt to satellite links, and the fragment size is dynamically adjusted to cope with link quality fluctuations and reduce transmission lag; the key fragment priority retransmission design focuses on the important information of the target object, so that the receiving end can quickly obtain the core content and ensure that the core value of the image is available first. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating the workflow of the image transmission method for geostationary orbit satellites provided by the present invention.

[0018] Figure 2 This invention provides a schematic diagram of an image transmission system suitable for geostationary orbit satellites.

[0019] Figure 3 This is an example of an image before intelligent cropping.

[0020] Figure 4 This is an example of an image after intelligent cropping. Detailed Implementation

[0021] To make the objectives, technical solutions, and effects of this invention clearer and more explicit, the invention is further described in detail below. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of the invention.

[0022] Please refer to the following first. Figure 1 The image transmission method applicable to geostationary orbit satellites includes the following steps: (1) The transmitting end acquires images; (2) Detect the target object in the acquired image and determine the edge coordinates of the target object; (3) Intelligent cropping of the image based on the edge coordinates of the target object, cropping out the invalid area outside the edge coordinates, and preserving the complete shape of the target object; (4) The cropped image is encoded using the HEIF encoding standard; (5) Dynamically adjust the data fragment size according to the current link quality parameters, divide the image data into multiple fragments, and add identification information to each fragment; (6) The sending end transmits the fragmented data to the receiving end via geostationary orbit satellite using the UDP protocol; (7) The receiving end receives the fragmented data, reassembles the fragments according to the identification information, and generates complete image data; if the receiving end detects lost fragmented data, it sends a list of lost fragments back to the sending end. (8) The sending end retransmits the lost fragment data according to the lost fragment list.

[0023] In step (1) above, the sending end refers to a UDP client capable of connecting to a geostationary orbit satellite. The sending end acquires images through an image acquisition module, which is a dedicated imaging device adapted to geostationary orbit satellite transmission scenarios. It must meet the requirements of high dynamic range, low noise, and environmental adaptability, including but not limited to spaceborne remote sensing cameras, ground-based high-definition monitoring cameras, and UAV aerial photography cameras. It must support the acquisition of static images or dynamic sequence images, and the acquisition resolution is configurable (ranging from 1280×720 to 8192×4320 pixels) to adapt to the needs of different transmission scenarios. For example, high resolution is required for remote sensing monitoring, while low resolution is used for rapid transmission in emergency communication.

[0024] Furthermore, in step (2), the YOLO algorithm is used for target detection. The purpose of using the YOLO algorithm is to adapt to the characteristics of "high latency and limited bandwidth" of satellite channels, while meeting the requirements of "fast speed and high accuracy" for target detection.

[0025] Furthermore, in step (2), if there are multiple targets in the image, each target is cropped separately without merging or splicing. The significance of this limitation is to maximize the reduction of invalid data. Due to the limited bandwidth of satellite channels, if multiple targets are merged and cropped (e.g., using a large rectangle to cover all targets), invalid areas between targets will be retained, resulting in a large amount of data after cropping, which will occupy the limited bandwidth. "Separate cropping" can remove the invalid areas around each target individually, retaining only the complete shape of a single target. Compared with merging and cropping, it can further reduce invalid areas (i.e., reduce redundant data), directly reducing the amount of data in subsequent HEIF encoding and fragmented transmission.

[0026] Further, in step (3), intelligent cropping is performed on the edge-free buffer area; after cropping, if the size of the effective area is too large, its resolution is reduced proportionally. The "edge-free buffer area" refers to pixel-level precise cropping based on the target object edge coordinates (x1, y1, x2, y2) determined in step (2), retaining only the target object pixels within the coordinate range, without adding extra edge buffer pixels. The (x1, y1, x2, y2) are the target object edge coordinates determined after detecting the target object using the YOLO algorithm in step (2), used to define the range of intelligent cropping. Specifically: the upper left corner of the original image before cropping is taken as the origin (0, 0), the horizontal direction to the right is the positive x-axis direction, and the vertical direction downward is the positive y-axis direction; (x1, y1) are the upper left corner pixel coordinates of the target object edge, and (x2, y2) are the lower right corner pixel coordinates of the target object edge. For ease of understanding, Figure 3 An example image before cropping is provided. Figure 4 Examples of corresponding intelligently cropped images are provided (in practical applications, these are color images; the attached images are simplified to black and white). (See attached...) Figure 3 and 4 As can be seen, there is no edge buffer area after cropping.

[0027] Furthermore, in step (3), the necessary metadata of the image is retained after intelligent cropping. The necessary metadata includes camera model, shooting time, and shooting location. "Shooting time + location" is matched with GIS to realize spatiotemporal analysis of the target; in emergency communication, "camera model" determines the imaging accuracy, and "shooting time" restores the event sequence to ensure the practical value of the image. In addition, the above-mentioned necessary metadata also facilitates problem tracing. For example, when packet loss / distortion occurs in satellite transmission, the "camera model" can be used to check equipment failure, "shooting time" can be used to check transmission delay, and "location" can be used to verify link matching to quickly locate the problem. In addition, the necessary metadata can also avoid secondary collection. Satellite collection resources are costly and time is limited. Retaining metadata eliminates the need to reschedule equipment for supplementary collection, reducing resource consumption and time costs.

[0028] Furthermore, in step (4), HEIF encoding uses quantization parameters for encoding control, and the quantization parameters are set to 30~40. The quantization parameter (QP value) reflects the spatial detail compression. The smaller the QP value, the more details are preserved; the larger the QP value, the more details are lost, and the lower the image clarity. A QP value of 30~40 is in a medium range, which can compress the amount of encoded data of the cropped image by 30%~50%, adapting to the limited bandwidth of the satellite. Since the sending end is a UDP client, the computing power is limited. A QP value of 30~40 does not require complex high-precision encoding calculations, which can reduce encoding time and the hardware load of the sending end, ensuring the overall real-time transmission performance. Furthermore, in step (5), the link quality parameters include: Signal-to-noise ratio (SNR) reflects the ratio of effective signal to background noise in a satellite communication link. Bit error rate (BER) refers to the ratio of the number of erroneous bits to the total number of bits transmitted during transmission. Link bandwidth utilization refers to the proportion of the actual bandwidth currently in use to the total available bandwidth of the satellite channel; Packet loss rate refers to the ratio of the number of fragments that were not successfully received to the total number of fragments sent.

[0029] When link quality degrades, the fragment size is reduced to 256 bytes; when link quality recovers, the fragment size returns to 512 bytes. In practical applications, satellite channels are affected by rain attenuation, ionospheric scintillation, etc. When link quality degrades (e.g., reduced signal-to-noise ratio, increased bit error rate), the large 512-byte fragments are easily lost due to channel interference, and retransmission requires retransmitting the entire large fragment, increasing the amount of retransmitted data and latency. Reducing the fragment size to 256 bytes results in shorter transmission time for smaller fragments and less exposure to unstable channels, thus reducing the probability of single fragment interference. Even if packets are lost, only 256-byte smaller fragments need to be retransmitted, significantly reducing the retransmission load and addressing the pain point of unstable satellite channels. When the link quality is restored and the channel transmission stability is enhanced, a large fragment of 512 bytes is used. Compared with a small fragment of 256 bytes, this reduces the total number of fragments, thereby reducing the total overhead of fragment identification information (sequence number, check bit, etc.). At the same time, it reduces the computational load of fragmentation at the sending end and reassembly at the receiving end, improves the overall transmission efficiency, and makes full use of the restored channel bandwidth resources.

[0030] Further, in step (5), the identification information includes fragment sequence number, total number of fragments, check bit, and timestamp. The "fragment sequence number" clarifies the position of each fragment in the complete data; the "total number of fragments" clearly tells the receiving end the total number of fragments to be received; the combination of "fragment sequence number" and "total number of fragments" allows the receiving end to quickly identify missing fragments (e.g., if the total number of fragments is 100, receiving fragments with sequence numbers 1-98 indicates missing fragments 99 and 100), and accurately reassemble the complete image data according to the sequence number, avoiding reassembly failure due to out-of-order delivery. The check bit performs verification calculations on the original data of each fragment, quickly determining whether the fragment has been damaged during transmission. For example, the receiving end compares the "check bit of the received fragment" with the "recalculated check bit"; if they are inconsistent, the fragment is determined to be invalid and needs to be added to the lost data list for retransmission. The timestamp is used to record the sending time of the fragment. The receiving end can judge the transmission time of the fragment by the current time and the timestamp. If the transmission time of the fragment exceeds the normal range (for example, more than 5 seconds compared to the preset threshold), it is determined to be a failed fragment with no real-time application value, and the receiving end is directly requested to retransmit the latest fragment.

[0031] Furthermore, in step (8), when the transmitting end retransmits a lost fragment, it prioritizes retransmitting the key fragment, which refers to a fragment containing important information about the target object. In satellite image applications (such as remote sensing and emergency response), the important information about the target object is the core of analysis and decision-making. Prioritizing the retransmission of key fragments containing this information can avoid non-key fragments occupying limited bandwidth, reduce the retransmission waiting time of core information, and allow the receiving end to quickly obtain the core content of the target object even when some non-key fragments have not arrived (for example, in an emergency scenario, the key features of the rescue vehicle are seen first), avoiding information delays caused by waiting for all fragments to be retransmitted. It can be seen that compared with indiscriminate retransmission, prioritizing the retransmission of key fragments can reduce the retransmission load of the transmitting end and the satellite channel occupation, achieving the core goal of "efficient transmission" of the system.

[0032] Accordingly, this invention provides an image transmission system suitable for geostationary orbit satellites, including a transmitter and a receiver. Unlike existing technologies, the transmitter is a UDP client capable of connecting to geostationary orbit satellites, used to execute steps (1) to (6) and (8) described above; the receiver is a UDP server connected to the terrestrial internet, used to execute step (7) of the method described above. Figure 2 As shown, the transmitting end consists of a satellite communication module and an image acquisition module, responsible for acquiring images and transmitting the image data to a geostationary orbit satellite via the satellite communication module. The geostationary orbit satellite acts as a relay, establishing a communication bridge between the transmitting end and the ground gateway to achieve data relay forwarding; after receiving the data from the satellite, the ground gateway transmits it to the receiving end via a data network. The receiving end includes a network adapter for a data access module, a CPU, a hard drive, etc. It receives data through the data network, the CPU is responsible for data processing, and the hard drive is used to store the received images and other data, completing the image reception and subsequent application stages.

[0033] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention in any way. It is understood that those skilled in the art can make equivalent substitutions or modifications based on the technical solution and inventive concept of the present invention, and all such modifications or substitutions should fall within the protection scope of the present invention.

Claims

1. An image transmission method suitable for geostationary orbit satellites, characterized in that, Includes the following steps: (1) The transmitting end acquires images; (2) Detect the target object in the acquired image and determine the edge coordinates of the target object; (3) Intelligent cropping of the image based on the edge coordinates of the target object, cropping out the invalid area outside the edge coordinates, and preserving the complete shape of the target object; (4) The cropped image is encoded using the HEIF encoding standard; (5) Dynamically adjust the data fragment size according to the current link quality parameters, divide the image data into multiple fragments, and add identification information to each fragment; (6) The sending end transmits the fragmented data to the receiving end via geostationary orbit satellite using the UDP protocol; (7) The receiving end receives the fragmented data, reassembles the fragments according to the identification information, and generates complete image data; If the receiving end detects lost fragment data, it sends a list of lost fragments back to the sending end. (8) The sending end retransmits the lost fragment data according to the lost fragment list.

2. The image transmission method for geostationary orbit satellites according to claim 1, characterized in that, In step (2), the YOLO algorithm is used to detect the target object.

3. The image transmission method for geostationary orbit satellites according to claim 1, characterized in that, In step (2), if there are multiple objects in the image, each object is cropped separately without merging or splicing.

4. The image transmission method for geostationary orbit satellites according to claim 1, characterized in that, In step (3), the edge-free buffer area is intelligently cropped; after cropping, if the size of the effective area is too large, its resolution is reduced proportionally.

5. The image transmission method for geostationary orbit satellites according to claim 1, characterized in that, In step (3), the necessary metadata of the image is retained after intelligent cropping, including camera model, shooting time, and shooting location.

6. The image transmission method for geostationary orbit satellites according to claim 1, characterized in that, In step (4), HEIF encoding is controlled by quantization parameters, which are set to 30~40.

7. The image transmission method for geostationary orbit satellites according to claim 1, characterized in that, In step (5), when the link quality degrades, the fragment size is reduced to 256 bytes; when the link quality recovers, the fragment size is restored to 512 bytes.

8. The image transmission method for geostationary orbit satellites according to claim 1, characterized in that, In step (5), the identification information includes the fragment sequence number, the total number of fragments, the check bit, and the timestamp.

9. The image transmission method for geostationary orbit satellites according to claim 1, characterized in that, In step (8), when the sending end retransmits a lost fragment, it prioritizes retransmitting the key fragment, which refers to a fragment containing important information about the target object.

10. An image transmission system suitable for geostationary orbit satellites, characterized in that, Including the sender and receiver, The sending end is a UDP client capable of connecting to geostationary orbit satellites, used to execute steps (1) to (6) and step (8) in the method described in claim 1. The receiving end is a UDP server connected to the terrestrial Internet, used to execute step (7) in the method described in claim 1.

Images