A Focus and Focus Auxiliary Judgment Method Based on Star Point Image Partitioning

By partitioning and stitching together astronomical photographic images, the problems of cumbersome operation and information loss in the traditional focusing process are solved, enabling fast and accurate global focus judgment and adapting to different devices and scenarios.

CN123118696APending Publication Date: 2026-11-17HANGZHOU INDEX STAR TECHNOLOGY CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611323857.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-28
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

In astrophotography, the traditional manual focusing process involves many steps, is inefficient in judgment, makes it difficult to observe the overall focus status at the same time, and causes information loss due to the movement of stars.

Method used

By partitioning, locating, cropping, and stitching the full-frame star image, and combining a safety duration threshold and a periodic repositioning mechanism, a global focus judgment map is generated. Users can observe the focus status of multiple areas in the same view, and reduce the risk of repeated positioning calculations and information loss.

Benefits of technology

It enables rapid and effective determination of global focus status, reduces data transmission and display burden, improves the efficiency and accuracy of the focusing process, and adapts to different sensors and scenarios.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123118696A_ABST
    Figure CN123118696A_ABST
Patent Text Reader

Abstract

This invention discloses a focus-assistance judgment method based on star point image partitioning. The method involves partitioning a full-frame star point image, locating star points, cropping local areas, and stitching the images together. During the initial run, a safety time threshold is introduced. Within this threshold, cropping is performed directly based on existing coordinates; after the threshold is exceeded, the star point coordinates are repositioned. This invention can present focus information for multiple regions of a full-frame image with a small data volume and high display efficiency during focusing. It also avoids requiring users to repeatedly zoom, drag, and switch between viewing different star point regions in the original large image. Furthermore, it balances the efficiency of star point coordinate positioning and the effectiveness of local image cropping even when star points move over time.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the fields of optical imaging, image sensor imaging, astrophotography, and focus-assisted display, and specifically to a focus-alignment auxiliary judgment method based on star point image partitioning. Background Technology

[0002] In low-light optical imaging scenarios such as astrophotography, users often need to judge the accuracy of focus based on the shape of stars. In traditional manual focusing, users usually need to repeatedly zoom in on local areas in a large original image, such as checking the shape of stars in the center or corners, then zoom out, move the field of view, and continue to zoom in on other areas. This process involves many steps, is inefficient, and makes it difficult to observe the focus status of multiple positions in the global field of view simultaneously in the same view.

[0003] For high-resolution image sensors, directly transmitting, displaying, and repeatedly zooming in on the entire image incurs a high data transmission and interaction burden. To determine global focus, users need to examine details in multiple dispersed areas, but the information truly useful for focus determination is often concentrated in local areas near individual stars. Therefore, a method is needed to automatically extract multiple representative local images of stars from the full image and display them centrally, allowing users to quickly determine global focus without repeatedly zooming and switching views.

[0004] Furthermore, in scenarios without an equatorial mount or without precise tracking, star points can move within the image due to factors such as the Earth's rotation. If the coordinates of the star points obtained from the initial localization are continuously used for cropping, the star points may move out of the cropping frame over time, resulting in loss of focus information. Therefore, a coordinate reuse and periodic relocalization mechanism is needed to reduce the computational load of repeated star point localization while preventing star points from moving out of the local cropping area.

[0005] Therefore, this invention proposes a focus and focus assistance method based on star point image partitioning. By partitioning the full-frame star point image, locating star points, cropping local areas, and stitching them together, the method assists users in judging the focus and focus status of the star point image. Summary of the Invention

[0006] The purpose of this invention is to propose a focus and focus assistance method based on star point image partitioning to overcome the shortcomings of existing technologies.

[0007] The method described in this invention is specifically as follows:

[0008] First, the full-frame image output by the optical imaging system is acquired and divided into N × N partitions. Within each partition, star points are located, identifying bright stars or target stars. Local images of preset sizes are then cropped using the target star's position as the center or reference point. These local images are then stitched together to form a global focus assessment image. Using this global focus assessment image, users can observe the star shapes from different regions of the original full-frame image on the same display screen, thus determining the focus status of the current star image.

[0009] Furthermore, in continuous display or multi-frame refresh scenarios, the host computer calculates a safe time threshold for the target star to move out of the current cropping frame during the first run. Based on parameters such as lens or telescope focal length, image sensor pixel size, cropping frame width and height, Earth's rotation parameters, and the shooting location, the safe time threshold is estimated using the most unfavorable movement speed, with a safety margin added. Before the safe time threshold expires, the system does not need to repeatedly search for the brightest star every frame; instead, it directly crops the corresponding local area based on the recorded coordinates and generates a stitched image. When the timer reaches the safe time threshold, the system re-acquires the current image and re-executes star positioning within the partition, updating the target star coordinates in each partition to prevent the star from moving out of the cropping frame.

[0010] Furthermore, the optical imaging system acquires a current full-frame image via an image sensor, which is derived from an astrophotography camera, a telescope system with an image sensor, an interchangeable lens imaging system, or other optical imaging devices capable of outputting star images.

[0011] Furthermore, the host computer divides the full-frame image into N parts horizontally and vertically, forming N × N partitions, based on imaging resolution, display screen size, target scene, and processing capabilities. Each partition can be an area of ​​equal size, or it can be adaptively divided according to the sensor's effective imaging area, display requirements, or edge distortion.

[0012] Furthermore, star point localization is achieved using XY peak finding, brightness projection, local extremum search, threshold screening, or other star point detection methods.

[0013] Furthermore, the host computer uses the coordinates of the target star as the center or reference point to crop a local image around the target star from the original full-frame image. The size of a single local image is determined based on the final stitched image resolution, display size, star size, focal length, pixel size, or user-set parameters.

[0014] Furthermore, the host computer stitches together multiple local images cropped from each partition into a global focus judgment image according to a preset order. The stitching order corresponds to the spatial location of the original partitions, allowing users to intuitively judge the focus status of stars in different areas of the full image from the stitched image. Since the stitched image only contains local information near the stars in each area, its data volume is much smaller than that of the complete original image, while the focus information is more concentrated.

[0015] Furthermore, users can determine whether the image is in focus by judging the shape of the stars in the image through global focus. For example, when the stars in different zones are all small, clear, or meet the preset visual standards, the focus is considered to be good; when the stars in certain areas are diffused, have ghosting, or have abnormal shapes, the user can be prompted to continue adjusting the focus or check for problems such as image field flatness and tilt.

[0016] The method described in this invention partitions, locates, locally crops, and stitches together the full-frame star image. At the same time, a safe duration threshold is introduced during the first run. Within the safe duration threshold, the image is directly cropped based on the existing coordinates. After the safe duration threshold is exceeded, the star coordinates are relocated.

[0017] By concentrating the focus information of scattered star points in the full-frame image into a single global focus judgment map, users can quickly determine the focus status of multiple areas within the same view. Transmitting or displaying only multiple local cropped images and their stitched results reduces reliance on the transmission and display of the complete high-resolution image, improving the response efficiency of the focusing process. Covering different image field locations such as the center, edges, and corners through an N × N partitioning method makes the global focus information more comprehensive. By using a safe duration threshold and a periodic relocation mechanism, repeated star point positioning calculations are reduced in scenarios where star points move, while also reducing the risk of information loss caused by star points moving out of the cropping frame. The number of partitions and cropping size are configurable, making it easy to adapt to different sensor resolutions, focal lengths, display devices, and usage scenarios.

[0018] It can present focus information of multiple areas of the full-frame image with a small amount of data and high display efficiency during focusing; at the same time, it avoids users having to zoom, drag and switch between different areas of stars in the original large image multiple times; and it balances the efficiency of star coordinate positioning and the effectiveness of local image cropping when stars move over time. Attached Figure Description

[0019] Figure 1 This is a flowchart of the method described in this invention;

[0020] Figure 2 This is a global focus determination diagram in the embodiment. Detailed Implementation

[0021] The present invention will be further described below with reference to the accompanying drawings.

[0022] In this embodiment, N × N is set to 5 × 5, that is, the entire image is divided into 25 equal regions, and the brightest star in each region is selected as the target star point.

[0023] In practice, the host computer can set the N value based on the display screen size, original image resolution, or computing resources. For example, a 3 × 3 partition can be used when the display area is small, while a 5 × 5 or higher number of partitions can be used for a more comprehensive global inspection.

[0024] like Figure 1 As shown, a focus-alignment auxiliary judgment method based on star point image partitioning is as follows:

[0025] In continuous display or multi-frame refresh scenarios, the host computer calculates a safe time threshold for the target star to move out of the current cropping frame during the first run. Based on parameters such as lens or telescope focal length, image sensor pixel size, cropping frame width and height, Earth's rotation parameters, and the shooting location, the safe time threshold is estimated using the most unfavorable movement speed, with a safety margin added. Before the safe time threshold expires, the system does not need to repeatedly search for the brightest star every frame; instead, it directly crops the corresponding local area based on the recorded coordinates and generates a stitched image. When the timer reaches the safe time threshold, the system re-acquires the current image and re-executes star localization within the partition, updating the target star coordinates in each partition to prevent the star from moving out of the cropping frame.

[0026] In scenarios without an equatorial mount or where star points are significantly shifting, the host computer first calculates a safe duration threshold based on focal length, pixel size, cropping frame size, Earth rotation parameters, and geographical location. Within the threshold, star point coordinates are reused; once the threshold is exceeded, star point positioning is re-executed, thereby reducing redundant peak-finding calculations and preventing star points from moving out of the cropping frame.

[0027] In this embodiment, the host computer acquires the full-frame image output by the optical imaging system, divides the full-frame image into N × N (5 × 5) partitions, and then divides the full-frame image into 25 equal regions. Within each region, the brightest star is found, and small images of the same size are cropped centered on that star. These small images are then stitched together in a 5 × 5 spatial order to form a global focus determination image (e.g., ...). Figure 2 (As shown). Users can observe the focus of stars in the center, edges, and corners of a single image.

[0028] In actual operation, the host computer can deduce the size of a single local cropped image based on the resolution of the final displayed image, or it can determine the cropping size based on the star point radius, focal length, or pixel size.

Claims

1. A focus and focus assist judgment method based on star point image partitioning, characterized in that, Specifically: First, the full-frame image output by the optical imaging system is acquired and divided into N × N partitions. Within each partition, star points are located to identify bright stars or target stars. Local images of a preset size are cropped using the position of the target star point in each partition as the center or reference position. Then, multiple local images are stitched together to form a global focus judgment image. Through this global focus judgment image, the user can observe the star point shapes from different areas of the original full-frame image on the same display screen, thereby determining the focus status of the current star point image.

2. The focus and focus assist judgment method based on star point image partitioning as described in claim 1, characterized in that, In continuous display or multi-frame refresh scenarios, the host computer calculates the safe time threshold for the target star to move out of the current cropping frame during the first run; it estimates the safe time threshold based on the most unfavorable moving speed; before the safe time threshold expires, the system does not need to repeatedly search for the brightest star every frame, but directly crops the corresponding local area based on the recorded coordinates and generates a stitched image; when the timer reaches the safe time threshold, the system reacquires the current image and re-executes the star positioning within the partition, updates the coordinates of the target star in each partition, and prevents the star from leaving the cropping frame due to movement.

3. The focus and focus assist judgment method based on star point image partitioning as described in claim 1, characterized in that, The optical imaging system acquires a current full-frame image through an image sensor, which comes from an optical imaging device capable of outputting star images.

4. The focus and focus assist judgment method based on star point image partitioning as described in claim 1, characterized in that, Star point localization is achieved using any one or more of the following methods: XY peak finding, brightness projection, local extremum search, or threshold filtering.

5. The focus and focus assist judgment method based on star point image partitioning as described in claim 1, characterized in that, The host computer uses the coordinates of the target star as the center or reference point to crop a local image around the target star from the original full-frame image; the size of a single local image is determined based on the resolution of the final stitched image, the screen size, the size of the star, the focal length, the pixel size, or user-set parameters.

6. The focus and focus assist judgment method based on star point image partitioning as described in claim 1, characterized in that, The host computer will stitch together multiple local images cropped from each partition into a global focus judgment image in a preset order; the stitching order corresponds to the spatial position of the original partition, allowing users to intuitively judge the focus status of stars in different areas of the full image from the stitched image.

7. The focus and focus assist judgment method based on star point image partitioning as described in claim 1, characterized in that, Users can judge whether the image is in focus by observing the shape of the stars in the image through global focus. When the stars in different areas are small, clear, or meet the preset visual standards, the focus is considered to be good. When the stars in some areas are diffused, have ghosting, or have abnormal shapes, the user can be prompted to continue adjusting the focus or check the flatness and tilt of the image field.