A detection display processing system for a portable underwater apparatus

Abstract

The application relates to a detection display processing system of an underwater portable mechanism, which comprises an fisheye image detection module, which is used for cooperating with a camera control module to provide an fisheye image detection function; the fisheye image detection function specifically comprises the following steps: acquiring multiple underwater images in real time, and performing edge detection on the underwater images to obtain edge images; performing straight line detection on the edge images to obtain a straight line segment set; taking the geometric center of the edge images as an origin, dividing the image plane into multiple concentric ring regions, and calculating the bending degree of the straight line segments located in the concentric ring regions; judging whether each underwater image is an fisheye image according to the bending degree; and an fisheye image conversion module, which is used for cooperating with the fisheye image detection module to realize fisheye image conversion and obtain multiple normal view angle images. Compared with the prior art, the application has the advantages of high image recognition and conversion precision.

Description

Technical Field

[0001] This invention relates to the fields of image processing and underwater detection technology, and in particular to a detection, display, and processing system for a portable underwater mechanism. Specifically, this system is a portable image detection, display, and processing system integrated into a mobile smart terminal for use in underwater environments. This invention is particularly targeted at underwater operation scenarios using standard UVC protocol cameras. Through hardware and software co-design, it achieves integrated functions of real-time underwater image acquisition, visual enhancement, intelligent analysis (including fisheye image stitching), and data management. Background Technology

[0002] With the increasing frequency of underwater resource exploration, engineering inspection, marine scientific research, and underwater recreational activities, the demand for portable, efficient, and low-cost underwater observation and recording tools has grown dramatically. Traditional underwater observation solutions mainly rely on two types of equipment: one is closed-circuit television systems carried by professional-grade underwater robots or manned submersibles, which are expensive, complex to deploy, and only suitable for large-scale projects; the other is consumer-grade waterproof cameras or action cameras, which, while portable, lack professional data processing and analysis capabilities, especially in underwater image enhancement and wide-field image acquisition. In practical applications, underwater operations often use cameras with fisheye lenses to obtain a wide field of view. However, while fisheye lenses offer the advantage of ultra-wide angles, they also introduce severe barrel distortion, resulting in distorted images that greatly affect the intuitiveness of observation and the accuracy of subsequent quantitative analysis. In existing technologies, the correction of fisheye images and the stitching of ordinary images are two relatively independent processes, lacking an intelligent system that can automatically identify image attributes (whether it is a fisheye image) and execute the corresponding processing flow. Users typically need to manually correct and stitch photos using professional software, a cumbersome process that cannot meet the needs of real-time or near-real-time processing on-site. Furthermore, the complex underwater lighting conditions present severe attenuation, scattering, and color cast issues (red light is absorbed first underwater), resulting in directly acquired images generally exhibiting low contrast, color distortion, and blurred details. While some mobile image apps with simple filter functions exist, these filters are mostly designed for terrestrial environments and lack adaptive optimization algorithms for underwater optical characteristics, limiting their processing effectiveness. At the same time, the standardized management of on-site data acquisition is a significant challenge. Image and video files acquired through traditional methods have inconsistent naming conventions and are difficult to accurately associate with shooting time and geographical location information, creating substantial difficulties for subsequent data processing, analysis, and archiving.

[0003] Chinese invention patent CN202410688912.6 discloses an "underwater image enhancement method, storage medium, and computer program product." While it effectively improves color cast and blurriness in underwater images through a deep learning model, it lacks a collaborative design with dedicated underwater imaging hardware and cannot achieve a real-time processing pipeline from image acquisition to display.

[0004] Patent document 2: CN202210757566.3 discloses "An underwater image enhancement method based on multi-domain information fusion". This method improves the realism and generalization ability of the enhancement effect, but its technical path relies heavily on complex deep learning models, making it difficult to achieve real-time or near-real-time operation on mobile smart terminals with limited computing power, and thus failing to meet the needs of on-site real-time analysis.

[0005] Patent document 3, CN202211629853.2, discloses "a fisheye image correction method and system based on geometric model and effective region extraction algorithm". This method is mainly aimed at ideal land or traffic monitoring scenarios, but it does not consider the interference of complex underwater optical characteristics on distortion correction accuracy and subsequent image stitching stability. When directly applied to underwater environments, the effect and robustness will be significantly reduced.

[0006] In summary, existing technologies have the following limitations: they are either limited to pure software enhancement and lack hardware collaboration; they focus on hardware integration of heavy platforms while neglecting intelligent processing; or the proposed correction algorithms are not applicable to complex underwater environments. Currently, there is an urgent need for a portable underwater detection, display, and processing system that integrates multi-sensor synchronization, real-time underwater image enhancement and correction, image stitching, and standardized data management to fill the gaps in existing technologies. Summary of the Invention

[0007] The purpose of this invention is to overcome the shortcomings of the prior art and provide a detection, display and processing system for an underwater portable mechanism.

[0008] The objective of this invention can be achieved through the following technical solutions: A detection, display, and processing system for an underwater portable mechanism, comprising: The camera control module is used to detect, connect to, and control the underwater camera, and dynamically display the frame rate, resolution, device information, and connection status; it is also used to respond to user commands for taking photos and recording videos. A fisheye image detection module is used to work in conjunction with the camera control module to provide fisheye image detection functionality. Specifically, the fisheye image detection functionality includes: acquiring multiple underwater images in real time and performing edge detection on the underwater images to obtain edge images; performing line detection on the edge images to obtain a set of line segments; dividing the image plane into multiple concentric annular regions with the geometric center of the edge images as the origin, and calculating the curvature of the line segments located within the concentric annular regions; and determining whether each underwater image is a fisheye image based on the curvature. The fisheye image conversion module works in conjunction with the fisheye image detection module to achieve fisheye image conversion. Specifically, it includes: constructing a distortion-free perspective image pixel grid based on the fisheye image; traversing each pixel of the distortion-free perspective image pixel grid and performing forward distortion calculation: calculating the distortion radial distance and azimuth angle corresponding to each pixel on the fisheye image, further calculating the corresponding pixel coordinates based on the distortion radial distance and azimuth angle, and assigning the corresponding pixel value to the corresponding pixel on the distortion-free perspective image pixel grid, until the traversal is completed, resulting in multiple conventional viewpoint images.

[0009] Furthermore, the system also includes an image stitching module, which is used to extract feature point information of each band in the overlapping area between consecutive frames of the multiple conventional viewpoint images, and perform weighted fusion to obtain fused band information; integrate the fused band information to obtain multiple preliminary panoramic images.

[0010] Furthermore, the multiple preliminary panoramic images are further stitched together to obtain an ultra-wide-angle panoramic image. The acquisition process of the ultra-wide-angle panoramic image specifically includes: performing spatiotemporal synchronization calibration on the multiple preliminary panoramic images to obtain multiple wide-angle images in a unified coordinate system; extracting feature points from the multiple wide-angle images using a feature point detection algorithm to obtain a feature point set; and performing feature matching between the multiple wide-angle images based on the feature point set to obtain a global feature point matching network; and constructing an error optimization model based on the global feature point matching network to optimize the stitching parameters. The error optimization model is iteratively optimized to obtain the optimal stitching parameter set. Based on the optimal stitching parameter set, the multiple wide-angle images are stitched together to obtain the final ultra-wide-angle panoramic image.

[0011] Furthermore, multiple wide-angle images under the unified coordinate system are reprojected and multi-band pixel fusion is performed to obtain the ultra-wide-angle panoramic image.

[0012] Furthermore, the ultra-wide-angle panoramic image is remapped to obtain an ultra-wide-angle panoramic image from a fisheye perspective.

[0013] Furthermore, the corresponding pixel coordinates are calculated based on the distorted radial distance and azimuth angle, using the following formula: in, and These are the pixel coordinates corresponding to the fisheye image. and Here are the normalized distortion coordinates for the fisheye image. The distorted radial distance, It is the azimuth angle. and Here, represents the normalized coordinates of pixels in the distortion-free perspective image, where u and v are the pixel coordinates of the distortion-free image. and The coordinates of the principal point in the image. and is the focal length, r is the normalized radial distance from a pixel in the distortion-free perspective image to the image center, and k1, k2, and k3 are the pre-calibrated radial distortion coefficients of the lens.

[0014] Furthermore, the system also includes a permission management module, which is used to request and obtain system camera, recording, file storage and location permissions, providing the basis for the operation of the functions; The data storage module is used to uniformly manage the saving and naming of media files, and to store photos or video files taken by underwater cameras in real time.

[0015] Furthermore, the system includes a flat panel mobile display device, multiple underwater cameras, an automatic lighting circuit, a depth sensor, a temperature sensor, and an underwater portable support mechanism; The flat panel mobile display device is connected to the multiple underwater cameras via a USB interface for video stream acquisition and control; each underwater camera is connected to the depth sensor, temperature sensor and automatic optical path completion to obtain depth information, temperature information and infrared light and illumination light compensation functions. The underwater portable support mechanism is used to connect the multiple underwater cameras, automatic lighting circuit, water depth sensor, and temperature sensor.

[0016] Furthermore, the system also includes an image enhancement module for implementing image enhancement functions; The image enhancement function specifically includes: allowing users to manually adjust the contrast and brightness of an image using a slider to uncover and enhance image details.

[0017] Furthermore, the data storage module is also used for: When acquiring images or videos, the system obtains the current geographic location information and system time, and associates and stores the geographic location information, system time, water depth information, and temperature information with the acquired images or videos.

[0018] Compared with the prior art, the present invention has the following advantages: (1) This invention proposes an efficient and accurate fisheye distortion correction and conversion process: by constructing a pixel grid of distortion-free perspective image as the target output framework, traversing each target pixel and calculating its corresponding distortion coordinates in the original fisheye image, and finally completing image reconstruction by assigning pixel values; by adopting a method that combines forward distortion calculation and reverse mapping, it ensures that the corrected image completely eliminates barrel distortion, restores the standard planar perspective geometry, provides high-quality input with geometric consistency for subsequent multi-image stitching, and significantly improves the accuracy and visualization effect of underwater scene reconstruction; Furthermore, by constructing an automatic fisheye image recognition mechanism based on geometric feature analysis, intelligent identification of underwater image types is achieved without manual annotation or preset parameters: concentric ring regions are divided with the image center as the origin, and the distribution pattern of the curvature of straight line segments in each region is quantitatively analyzed to accurately distinguish between fisheye distortion images and conventional perspective images; this recognition process is entirely based on the characteristics of the image content itself, and has the advantages of strong adaptability, high judgment accuracy, and no dependence on external markers, providing a reliable decision basis for subsequent targeted processing.

[0019] (2) This invention, through the design of an intelligent fisheye image stitching engine, realizes a complete process of automatic fisheye image recognition, distortion correction, feature matching, and panoramic stitching without manual intervention. This engine can perform high-precision registration and multi-band fusion of corrected multi-view images to generate seamless ultra-wide-angle underwater scenes, effectively solving the problems of insufficient field of view and image distortion in traditional underwater observation. Combined with the GPS positioning function of the mobile terminal and the fusion of information from multiple sensors such as water depth and temperature, the system realizes automatic association and standardized management of shooting data and spatiotemporal environmental information, forming a traceable and easily searchable underwater image database.

[0020] (3) The system of the present invention has the capability of underwater environment adaptive image enhancement. Through automatic supplementary lighting and real-time advanced filters in hardware and software collaboration, it effectively compensates for underwater light attenuation, scattering and color shift, and improves image clarity and color authenticity. Its overall design takes into account processing efficiency and user experience, and supports the completion of the entire process of real-time preview, parameter adjustment, image acquisition, post-processing and data archiving within a single mobile application. It is suitable for a variety of professional scenarios such as underwater surveying, engineering inspection, ecological survey, and archaeological recording, and provides an efficient, portable and comprehensive solution for underwater field operations. Attached Figure Description

[0021] Figure 1 This is a system flowchart of a detection, display, and processing system for an underwater portable mechanism provided in an embodiment of the present invention; Figure 2 This is a flowchart illustrating the operation of a detection, display, and processing system for an underwater portable mechanism provided in an embodiment of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0023] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0024] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0025] Example 1 like Figure 1 As shown, this embodiment provides a detection, display, and processing system for an underwater portable mechanism, including: The camera control module is used to detect, connect to, and control the underwater camera, and dynamically display the frame rate, resolution, device information, and connection status; it is also used to respond to user commands for taking photos and recording videos. The fisheye image detection module works in conjunction with the camera control module to provide fisheye image detection functionality. This functionality specifically includes: acquiring multiple underwater images in real time and performing edge detection on the underwater images to obtain edge images; performing line detection on the edge images to obtain a set of line segments; dividing the image plane into multiple concentric annular regions with the geometric center of the edge images as the origin, and calculating the curvature of the line segments within these concentric annular regions; and determining whether each underwater image is a fisheye image based on the degree of curvature. Fisheye Image Conversion Module: This module works in conjunction with the fisheye image detection module to achieve fisheye image conversion. Specifically, it includes: constructing a distortion-free perspective image pixel grid based on the fisheye image; traversing each pixel of the distortion-free perspective image pixel grid and performing forward distortion calculation: calculating the distortion radial distance and azimuth angle corresponding to each pixel in the fisheye image, further calculating the corresponding pixel coordinates based on the distortion radial distance and azimuth angle, and assigning the corresponding pixel value to the corresponding pixel in the distortion-free perspective image pixel grid, until the traversal is complete, resulting in multiple images from conventional perspectives.

[0026] Preferred, The system also includes an image stitching module, which is used to extract feature point information of each band in the overlapping area between consecutive frames of multiple conventional viewpoint images, and perform weighted fusion to obtain fused band information; integrate the fused band information to obtain multiple preliminary panoramic images.

[0027] Preferred, Spatiotemporal synchronization calibration is performed on multiple preliminary panoramic images to obtain multiple wide-view images in a unified coordinate system; feature points are extracted from multiple wide-view images using a feature point detection algorithm to obtain a feature point set; and feature matching is performed between multiple wide-view images based on the feature point set to obtain a global feature point matching network; based on the global feature point matching network, an error optimization model is constructed to optimize the stitching parameters. The error optimization model is iteratively optimized to obtain the optimal stitching parameter set. Based on the optimal stitching parameter set, multiple wide-angle images are stitched together to obtain the final ultra-wide-angle panoramic image.

[0028] Preferred, By reprojecting and multi-band pixel fusion of multiple wide-angle images under a unified coordinate system, an ultra-wide-angle panoramic image is obtained.

[0029] Preferred, Further remapping of the ultra-wide-angle panoramic image yields an ultra-wide-angle panoramic image from a fisheye perspective.

[0030] Preferred, The corresponding pixel coordinates are calculated based on the distorted radial distance and azimuth angle. The corresponding calculation formula is as follows: in, and These are the pixel coordinates corresponding to the fisheye image. and Here are the normalized distortion coordinates for the fisheye image. The distorted radial distance, It is the azimuth angle. and Here, represents the normalized coordinates of pixels in the distortion-free perspective image, where u and v are the pixel coordinates of the distortion-free image. and The coordinates of the principal point in the image. and is the focal length, r is the normalized radial distance from a pixel in the distortion-free perspective image to the image center, and k1, k2, and k3 are the pre-calibrated radial distortion coefficients of the lens.

[0031] Preferred, The distortion-free coordinates are calculated based on the pixel values ​​of the fisheye image using the following formula: The distortion-free radial distance r is obtained by solving the inverse distortion problem using an iterative method, and then: in, These are distortion-free coordinates.

[0032] The core of this system lies in addressing the problems of insufficient image clarity, limited target field of view, and lack of location information in underwater detection in turbid waters. The system features functions such as image clarity enhancement, fisheye lens image calibration, multi-camera image registration and stitching, and information data location. To address issues such as light attenuation, color distortion, and insufficient contrast in underwater environments, the system achieves simultaneous hardware and software enhancement of camera image information through an automatic supplemental lighting circuit and advanced image filter enhancement software on a flat panel display device. The system utilizes the GPS positioning system of the flat panel display device and information fusion from depth and temperature sensors to construct comprehensive location information, facilitating data management and geographic positioning. The flat panel display device enables comprehensive image processing, including automatic camera lens recognition, fisheye distortion image calibration, advanced image filter enhancement, single-camera image stitching, multi-camera image stitching, comprehensive location information processing, and data storage.

[0033] Specifically, The flat panel mobile display device connects to multiple underwater cameras via a USB interface for video stream acquisition and control; Each underwater camera is connected to an automatic illumination circuit, a depth sensor, and a temperature sensor to obtain water depth information, temperature information, and hardware infrared light and illumination light compensation functions. The underwater portable support mechanism is used to connect multiple underwater cameras, automatic lighting circuits, depth sensors, and temperature sensors to achieve the function of portable support. Flat panel mobile display devices are used to achieve comprehensive image processing functions, including automatic lens recognition of cameras, fisheye distortion image calibration, advanced image filter enhancement, single-camera image stitching, multi-camera field-of-view stitching, information data comprehensive positioning information processing, and data storage functions.

[0034] Specifically, Advanced image filter enhancement features are achieved through image processing algorithms built into the tablet mobile display device. The processed image significantly improves visual clarity while preserving details, effectively overcoming underwater light scattering and color cast issues.

[0035] The fisheye distortion image calibration function, based on pre-calibrated camera intrinsic parameters and distortion coefficients, uses a reverse mapping algorithm to reproject curved pixels in the fisheye image onto a planar coordinate system under an ideal pinhole camera model. The calibrated image eliminates barrel distortion, providing geometrically accurate input data for subsequent image stitching and significantly improving the quantitative level of underwater observation.

[0036] Preferred, The single-camera image stitching function analyzes the overlapping areas between consecutive frames taken by the same camera at different locations, extracts feature points for matching, and generates panoramic images through a multi-band fusion algorithm.

[0037] The single-camera image stitching function analyzes the overlapping areas between consecutive frames captured by the same camera at different locations, extracts feature points for matching, and then uses a multi-band fusion algorithm. This involves extracting the band features of each band from the aligned overlapping areas of adjacent frames, performing weighted average fusion on each band, and then integrating the fused band information. Gradient blurring is then applied to the edges of the overlapping areas to eliminate stitching seams, thus generating a panoramic image. The analysis of overlapping areas between consecutive frames captured by the same camera at different locations refers to locating and analyzing the repeatedly captured areas in two consecutive frames when the single camera moves with the underwater portable device, determining the location and overlap rate of these overlapping areas, and limiting the effective range for subsequent feature point matching.

[0038] Preferred, Multi-camera field-of-view stitching performs spatiotemporal synchronization calibration on images acquired by each camera, unifies the coordinate system by calibrating parameters, and then uses a global optimization strategy to fuse multi-view data to achieve complete reconstruction of ultra-wide-angle underwater scenes.

[0039] Multi-camera field-of-view stitching performs spatiotemporal synchronous calibration on images captured by each camera, unifies the coordinate system through calibration parameters, and then employs a global optimization strategy to fuse multi-view data. This approach takes into account the perspective data from all cameras and globally corrects errors in the multi-camera stitching process, ensuring that the final generated ultra-wide-angle underwater scene is free of local misalignment, geometrically consistent, and detailed. This meets the needs of multi-camera collaborative detection in portable underwater devices, achieving complete reconstruction of ultra-wide-angle underwater scenes. Specifically, the global optimization strategy involves extracting feature points from all camera images, establishing perspective correlations and constructing a global feature point matching network, building an error optimization model that incorporates various error factors, iteratively calculating the optimal solution, and simultaneously correcting perspective data and stitching parameters to eliminate local deviations, resulting in a unified, coherent, and distortion-free complete ultra-wide-angle scene.

[0040] Preferred, The integrated positioning system obtains latitude and longitude information through the GPS function of the tablet mobile display device, and then formats and splices it with the system time information of the tablet mobile display device, the depth information of the water depth sensor, and the temperature information of the temperature sensor.

[0041] This system combines a high-performance underwater USB camera with a customized Android application, forming a complete closed-loop solution from data acquisition and real-time processing to intelligent analysis and archiving management. Through deep hardware and software integration, this invention achieves the following technical functions: 1. Hardware Compatibility and Real-time Monitoring: The system is compatible with various underwater USB cameras that support the standard UVC protocol, enabling the detection, connection, and control of UVC-compliant underwater USB cameras. The software interface dynamically overlays key status parameters in real time, including video frame rate, current resolution, connected device ID, and connection status, providing users with comprehensive system monitoring.

[0042] 2. Underwater Environment Adaptive Image Processing: The system incorporates an image processing pipeline optimized for underwater environments. Firstly, during the preview and acquisition stages, the system integrates an underwater environment adaptive filter module. This module includes advanced image filters, allowing users to select preset options and adjust parameters in real-time based on water quality and lighting conditions to compensate for underwater optical attenuation, significantly enhancing the subjective visual effect of the image. Secondly, in the post-processing stage, image enhancement functions are provided, allowing users to non-linearly adjust the contrast and brightness of saved images to uncover image details.

[0043] 3. Intelligent Fisheye Image Stitching Engine: The system's integrated image stitching function can automatically detect whether the image to be processed was captured by a fisheye lens. Its workflow is as follows: The system first performs feature analysis on the image to determine its distortion type. If it is determined to be a normal perspective image, it directly uses a feature point matching algorithm for stitching. If it is identified as a fisheye image, it first calls the built-in fisheye distortion correction model to restore the image to a standard planar perspective image, then performs high-precision feature point matching and image fusion stitching, and finally remaps the stitched panoramic image to a fisheye perspective, thus obtaining a wide, seamless panoramic view while maintaining the consistency of the final output image's perspective. This intelligent process eliminates the tediousness of users manually selecting processing modes, greatly improving processing efficiency and ease of use.

[0044] Specifically, The system performs edge detection on the image under test to obtain edge images. Using Hough transform, it finds straight line segments in the output image and divides the image into multiple concentric annular regions with the geometric center of the image as the origin. For each straight line segment within a region, it calculates the actual curvature. If the actual curvature of each region satisfies the characteristic of radial distortion in fisheye lens images—"weak at the center, strong at the edges"—it is determined to be a fisheye image; otherwise, it is identified as a normal perspective image. If it is determined to be a normal perspective image, a feature point matching algorithm is directly used for stitching. If it is identified as a fisheye image, the built-in fisheye distortion correction model is invoked to restore the image to a standard planar perspective image. The model retrieves the core parameters pre-calibrated for the current lens, automatically generates a uniform-sized pixel grid of distortion-free perspective images based on the resolution of the input fisheye image, traverses the pixels, performs forward distortion calculation to obtain the radial distance of the distortion at each point, and records the azimuth angle. Then, through inverse mapping coordinate transformation, it obtains the positions of the distortion-free points and the corresponding distortion pixels in the fisheye image. Pixel values ​​from each position in the fisheye image are extracted and assigned to the corresponding points in the distortion-free image. After traversal, a standard planar perspective image is output. Then, high-precision feature point matching and image fusion stitching are performed. Finally, the stitched panoramic image is remapped to a fisheye view, thus obtaining a wide and seamless panoramic field of view while maintaining the consistency of the final output image's perspective. This intelligent process eliminates the tediousness of users manually selecting processing modes, greatly improving processing efficiency and ease of use.

[0045] 4. Standardized Data Acquisition and Management: When taking photos and videos, if the device network is available, the system will automatically acquire accurate GPS or network positioning information and automatically generate filenames with the shooting time and latitude / longitude coordinates according to a preset format. This naming method ensures that each piece of data has a unique and traceable time and space tag, greatly facilitating the archiving, retrieval, and analysis of underwater data.

[0046] 5. Smooth user interaction experience: The entire system adopts a modular design, with the main interface clearly divided into two major functional areas: "Camera Preview" and "Image Processing". Users can seamlessly and quickly switch between the two functional modules through intuitive buttons without having to exit the application or restart the service, ensuring the continuity and efficiency of the operation process.

[0047] The advantages of this invention lie in its combination of professional underwater image processing capabilities with the portability of mobile devices, providing a low-cost, high-efficiency, and comprehensive field solution. Its intelligent fisheye image processing workflow and standardized data management methods are particularly suitable for professional fields requiring real-time acquisition of high-quality, stitchable, and traceable image data, such as underwater surveying, engineering inspection, ecological surveys, and archaeological recording. This invention integrates image acquisition, real-time enhancement, and intelligent stitching, providing an efficient and portable hardware and software solution for underwater exploration, engineering inspection, and scientific research recording.

[0048] Example 2 like Figure 1 and Figure 2 As shown, this embodiment provides a specific detection, display, and processing system for an underwater portable mechanism. The system is primarily implemented using the Android Studio development environment, written in Kotlin / Java, and compatible with Android 15 and above operating systems. It includes: I. System Overall Structure and Operation Process like Figure 2 As shown, this system mainly consists of a tablet mobile display device (running the Android operating system) and multiple underwater sensing and acquisition units. Each acquisition unit includes an underwater USB camera compliant with the UVC protocol, and integrates a temperature sensor, a depth sensor, and an automatic illumination circuit, and is connected to the tablet device via a wired connection.

[0049] The overall operation of the system follows the principle of modular design and is mainly divided into four functional modules: 1. Permission Management Module: After the application starts, it first requests and obtains system permissions for camera, recording, file storage and precise location, providing the foundation for the operation of subsequent functions.

[0050] 2. Camera Control Module: Responsible for detecting, connecting to, and controlling the underwater USB camera. This module enables real-time preview of the video stream, dynamically displaying frame rate (FPS), resolution, device information, and connection status. Users can manually switch preset resolutions (including 1920×1080, 1600×1200, etc.) through the interface and adjust parameters such as brightness, contrast, and saturation in real time to compensate for underwater optical attenuation and optimize image quality. This module also responds to user commands for taking photos and recording videos.

[0051] 3. Image Processing Module: Real-time Filters and Post-processing This module integrates two core functions.

[0052] First, there's real-time image filtering and parameter adjustment. This function works in conjunction with the camera control module, allowing users to adjust the image's tone, sharpness, and other attributes in real time based on the underwater environment before capturing the image, achieving the best pre-processing results.

[0053] Secondly, it provides post-processing image functions. Users can select saved images from their local photo album for further processing. Specifically, this includes: (1) Image enhancement: Allows users to manually adjust the contrast and brightness of the image using a slider to uncover and enhance image details.

[0054] (2) Intelligent image stitching: This function can automatically identify whether the two input images are taken with a fisheye lens. If they are identified as ordinary images, feature point matching and stitching are performed directly; if they are fisheye images, the built-in geometric correction model is first called to restore the image to a normal perspective view, and then feature matching, image alignment and seamless fusion are performed. After the stitching is completed, the system will remap the panoramic image to a fisheye view to maintain visual consistency.

[0055] 4. Data Storage Module: Unified management of media file saving and naming. When taking photos or videos, this module calls the location service, combines precise latitude and longitude information with the system timestamp, automatically generates a standard filename in the format of "shooting time-latitude and longitude", and saves it to a designated directory in the device's photo album, achieving standardized data management.

[0056] II. Implementation Process of Key Functions 1. Camera initialization and parameter control After the system starts up, the camera control module initializes the video capture session according to the resolution parameters selected by the user. Once the camera is successfully connected, the real-time video stream is rendered to the preview interface. Users can adjust the parameters by sliding the sliders on the interface; these adjustments are sent to the camera hardware in real time or processed by the software before taking effect, allowing for instant adjustments to the image quality.

[0057] 2. Geolocation and Timestamp Synthesis: When a user triggers a photo or starts recording video, the data storage module immediately invokes the system's location service. This service prioritizes obtaining the device's most recent known location; if not, it initiates a rapid network or GPS location request. After obtaining valid latitude and longitude coordinates, they are combined with the current system time and formatted into a standardized filename according to predetermined rules.

[0058] 3. Media capture and preservation (1) Taking a picture: The system captures the image frame of the current preview screen, associates the image data with the aforementioned standard file name, encodes it into JPEG format, and saves it to the device's public album directory.

[0059] (2) Video recording: The system starts the video encoder to continuously capture video frames and mix them with the audio stream. The video frames are also encoded into MP4 format files with standard filenames containing spatiotemporal information and stored.

[0060] 4. Image processing function implementation 1. Image Enhancement: After selecting an image, users can adjust the contrast and brightness parameters using sliders. The system then performs global or local transformations on the image pixel data based on these parameter values, enhancing the image's visual detail and overall appearance.

[0061] 2. Intelligent Image Stitching: a. Feature Extraction: The system automatically analyzes the two selected images to be stitched, extracting key feature points. b. Fisheye Recognition and Correction: The system determines whether the image was captured with a fisheye lens. If so, it first applies a built-in geometric correction model to restore the image to a normal perspective. c. Feature Matching and Alignment: Feature points from different images are matched, and the optimal spatial transformation relationship is calculated to align the content of the two images. d. Fusion and Output: The aligned images are smoothly fused to eliminate seams. For fisheye images, the result is remapped back to the fisheye perspective after stitching to maintain visual consistency. Finally, a seamless panoramic image is generated.

[0062] III. User Operation Flow After launching the application, users first complete the necessary permission authorization. They then enter the main preview interface, where they can view underwater footage in real time, adjust acquisition parameters, and take photos or videos. All media files are automatically saved in a standardized format. When post-processing is required, users can click the switch button to enter the image processing interface, where they can select images for enhancement or stitching operations; the processed results can also be saved to the photo album. The entire process is completed within a single application, with seamless operation, eliminating the need to switch between different software.

[0063] The preferred embodiments of the present invention have been described in detail above. It should be understood that those skilled in the art can make numerous modifications and variations based on the concept of the present invention without creative effort. Therefore, all technical solutions that can be obtained by those skilled in the art based on the concept of the present invention through logical analysis, reasoning, or limited experimentation on the basis of existing technology should be within the scope of protection defined by the claims.

Claims

1. A detection, display, and processing system for an underwater portable mechanism, characterized in that, include: The camera control module is used to detect, connect to, and control the underwater camera, and dynamically display the frame rate, resolution, device information, and connection status. Used to respond to user commands to take photos and record videos; A fisheye image detection module is used to work in conjunction with the camera control module to provide fisheye image detection functionality. Specifically, the fisheye image detection functionality includes: acquiring multiple underwater images in real time and performing edge detection on the underwater images to obtain edge images; performing line detection on the edge images to obtain a set of line segments; dividing the image plane into multiple concentric annular regions with the geometric center of the edge images as the origin, and calculating the curvature of the line segments located within the concentric annular regions; and determining whether each underwater image is a fisheye image based on the curvature. The fisheye image conversion module works in conjunction with the fisheye image detection module to achieve fisheye image conversion. Specifically, it includes: constructing a distortion-free perspective image pixel grid based on the fisheye image; traversing each pixel of the distortion-free perspective image pixel grid and performing forward distortion calculation: calculating the distortion radial distance and azimuth angle corresponding to each pixel on the fisheye image, further calculating the corresponding pixel coordinates based on the distortion radial distance and azimuth angle, and assigning the corresponding pixel value to the corresponding pixel on the distortion-free perspective image pixel grid, until the traversal is completed, resulting in multiple conventional viewpoint images.

2. The detection, display, and processing system for an underwater portable mechanism according to claim 1, characterized in that, The system also includes an image stitching module, which is used to extract feature point information of each band in the overlapping area between consecutive frames of the multiple conventional viewpoint images, and perform weighted fusion to obtain fused band information; integrate the fused band information to obtain multiple preliminary panoramic images.

3. The detection, display, and processing system for an underwater portable mechanism according to claim 2, characterized in that, The multiple preliminary panoramic images are further stitched together to obtain an ultra-wide-angle panoramic image; The acquisition process of the ultra-wide-angle panoramic image specifically includes: firstly, performing spatiotemporal synchronization calibration on the multiple preliminary panoramic images to obtain multiple wide-angle images in a unified coordinate system; secondly, using a feature point detection algorithm to extract feature points from the multiple wide-angle images to obtain a feature point set, and based on the feature point set, performing feature matching between the multiple wide-angle images to obtain a global feature point matching network; and thirdly, based on the global feature point matching network, constructing an error optimization model to optimize the stitching parameters. The error optimization model is iteratively optimized to obtain the optimal stitching parameter set. Based on the optimal stitching parameter set, the multiple wide-angle images are stitched together to obtain the ultra-wide-angle panoramic image.

4. The detection, display, and processing system for an underwater portable mechanism according to claim 3, characterized in that, The ultra-wide-angle panoramic image is obtained by reprojecting and multi-band pixel fusion of multiple wide-angle images under the unified coordinate system.

5. The detection, display, and processing system for an underwater portable mechanism according to claim 4, characterized in that, The ultra-wide-angle panoramic image is further remapped to obtain an ultra-wide-angle panoramic image from a fisheye perspective.

6. The detection, display, and processing system for an underwater portable mechanism according to claim 1, characterized in that, The corresponding pixel coordinates are calculated based on the distorted radial distance and azimuth angle, and the corresponding calculation formula is as follows: in, and These are the pixel coordinates corresponding to the fisheye image. and Here are the normalized distortion coordinates for the fisheye image. The distorted radial distance, It is the azimuth angle. and Here, represents the normalized coordinates of pixels in the distortion-free perspective image, where u and v are the pixel coordinates of the distortion-free image. and The coordinates of the principal point in the image. and is the focal length, r is the normalized radial distance from a pixel in the distortion-free perspective image to the image center, and k1, k2, and k3 are the pre-calibrated radial distortion coefficients of the lens.

7. The detection, display, and processing system for an underwater portable mechanism according to claim 1, characterized in that, The system also includes a permission management module, which is used to request and obtain system camera, recording, file storage and location permissions, providing the basis for the operation of the functions; The data storage module is used to uniformly manage the saving and naming of media files, and to store photos or video files taken by underwater cameras in real time.

8. The detection, display, and processing system for an underwater portable mechanism according to claim 7, characterized in that, The system includes a flat panel mobile display device, multiple underwater cameras, an automatic fill light circuit, a water depth sensor, a temperature sensor, and an underwater portable support mechanism. The flat panel mobile display device is connected to the multiple underwater cameras via a USB interface for video stream acquisition and control; each underwater camera is connected to the depth sensor, temperature sensor and automatic optical path completion to obtain depth information, temperature information and infrared light and illumination light compensation functions. The underwater portable support mechanism is used to connect the multiple underwater cameras, automatic lighting circuit, water depth sensor, and temperature sensor.

9. The detection, display, and processing system for an underwater portable mechanism according to claim 7, characterized in that, The system also includes an image enhancement module for implementing image enhancement functions; The image enhancement function specifically includes: allowing users to manually adjust the contrast and brightness of an image using a slider to uncover and enhance image details.

10. The detection, display, and processing system for an underwater portable mechanism according to claim 7, characterized in that, The data storage module is also used for: When acquiring images or videos, the system obtains the current geographic location information and system time, and associates and stores the geographic location information, system time, water depth information, and temperature information with the acquired images or videos.

Images