Indoor vision positioning method and positioning system based on occlusion clearing algorithm

Abstract

The invention discloses an indoor visual positioning method and positioning system based on an occlusion clearing algorithm, and belongs to the technical field of occlusion image processing, image restoration and indoor positioning. The method is characterized by comprising the following steps: step 1, querying an image in a database, extracting feature information of the image in the database, and storing the feature information; step 2, shielding object detection is carried out, and shielding objects in all query images are removed; 3, performing image restoration by adopting the image restoration model, and restoring a query image; and step 4, indoor positioning is carried out. According to the indoor vision positioning method and positioning system based on the occlusion clearing algorithm, target detection, image restoration and feature matching technologies are combined, and the positioning accuracy and robustness are improved. According to the method and the device, the blocking objects in all the query images are detected and removed, so that the pixels of the blocking objects do not interfere with subsequent processing, and meanwhile, a clear target area is provided for an image recovery process.

Description

Technical Field

[0001] The present invention relates to the technical fields of occlusion image processing, image restoration and indoor positioning, and in particular to an indoor visual positioning method and a positioning system based on an occlusion removal algorithm. Background Art

[0002] Indoor positioning technology has important applications in modern smart homes, security monitoring, robot navigation and other fields. With the development of technology, vision-based positioning systems have gradually become a hot topic of research due to their low cost, easy installation, and rich information. These methods obtain environmental images through cameras and use image features for positioning and navigation. Compared with traditional methods based on sensors such as odometers and lidar, visual positioning systems can not only provide richer environmental information, but also achieve higher positioning accuracy.

[0003] However, due to the presence of a large number of obstructing objects in indoor environments, such as furniture and decorations, these obstructing objects will affect the camera's field of view, resulting in incomplete image information, which will greatly damage the positioning accuracy and may even cause the loss of key feature points. In this case, the traditional visual positioning algorithm cannot effectively process the occlusion information, resulting in increased positioning errors or even inability to locate. In addition, the occluded objects may also cause misjudgment of the visual algorithm, misidentifying the occluded objects as environmental features, further exacerbating the difficulty of positioning. Therefore, how to effectively process occluded images and restore complete environmental information has become an important research direction for improving indoor positioning accuracy. Summary of the invention

[0004] The technical problem to be solved by the present invention is: to overcome the shortcomings of the prior art and provide an indoor visual positioning method and positioning system based on an occlusion removal algorithm that combines target detection, image restoration and feature matching technology to improve positioning accuracy and robustness.

[0005] The technical solution adopted by the present invention to solve the technical problem is: the indoor visual positioning method based on the occlusion removal algorithm is characterized by comprising the following steps: Step 1, query the image in the database, extract the feature information of the database image, and save the feature information; Step 2: Detect occluders and remove all occluders in the query image. Step 3: Use the image restoration model to perform image restoration and restore the query image; Step 4: Perform indoor positioning.

[0006] Preferably, when executing step 2, it includes: constructing an occlusion image dataset; performing occlusion detection based on the occlusion image dataset; and resetting pixels to eliminate occlusions in the query image.

[0007] Preferably, when constructing the occluded image dataset: a number of objects are obtained in the coco128 dataset by matting, where the matting process involves separating the object of interest from the image to obtain an object image with a transparent background; then, mask images of these objects are generated by retaining the pixels of the object and setting the pixel values ​​of the remaining positions to (255, 255, 255); Load the images of the 7Scenes dataset; then, load the images of all occluded objects and their corresponding masks; randomly select an occluded object and its mask for each query and database image, and add random types and degrees of blur information to them, and then place the occluded object with blurred information at random positions on the image with random scales and rotation angles to generate occluded images. Finally, the occluded images generated one-to-one for all images constitute the occluded image dataset.

[0008] Preferably, the steps for performing occluder detection based on the occluded image dataset are: first read the bounding box information in the detection result, extract the pixel area of ​​the occluded object, load the YOLOv5 model pre-trained on the COCO dataset, detect the query image, generate a bounding box and category information containing the occluded object, and complete the detection of the occluded object.

[0009] Preferably, the process of pixel resetting to eliminate occluders in the query image is: resetting the pixels in the occluded object area to pure black pixels (0, 0, 0) to indicate the position of the occluded object; at the same time, generating a mask image corresponding to the query image, in which the pixels in the occluded object area are set to pure black pixels, and the pixels in other areas are set to pure white pixels (255, 255, 255).

[0010] Preferably, step 4 comprises the following steps: Step 4-1, extracting and saving feature information of the query image restored in step 3; Step 4-2, performing image retrieval; performing a neighbor retrieval of the query image using an image retrieval algorithm based on the database image feature information and the query image feature information to obtain a neighbor retrieval pair of the query image; Step 4-3, based on the database image feature information, the query image feature information and the nearest neighbor retrieval pair, use the LightGlue algorithm to perform feature matching and establish a 2D-2D matching relationship between the query image and the database image; Step 4-4, using the indoor visual positioning algorithm based on the essential matrix; based on the 2D-2D matching information, the five-point method in the RANSAC cycle is used to calculate the essential matrix of the image pair; then, through the epipolar constraint condition, the singular value decomposition and RANSAC algorithm are used, combined with the known absolute pose and essential matrix of the database image, to finally determine the positioning result.

[0011] Preferably, when executing step 1, firstly, a reference point is set in the indoor environment, and image data is collected at the reference point; secondly, the database image is extracted using the SuperPoint algorithm and the NetVLAD algorithm for feature extraction; finally, the extracted feature information and the corresponding coordinate information are stored to form a feature file.

[0012] Preferably, in step 3, an image restoration model MAT is used to perform image restoration.

[0013] An indoor visual positioning system based on an occlusion removal algorithm is characterized by comprising an offline unit and an online unit, wherein the offline unit comprises: Offline image acquisition module, used to query images in the database; The offline feature extraction module is connected to the offline image acquisition module and is used to extract feature information of the database image and save the feature information; Online units include: An online image acquisition module, used to obtain query images; The occlusion removal module is connected to the online image acquisition module and is used to detect occlusions and remove occlusions in all query images; The indoor positioning module is connected to the occlusion removal module and the offline feature extraction module and is used for indoor positioning.

[0014] Preferably, the indoor positioning module includes: A feature extraction submodule is used to obtain feature information of the restored query image; The image retrieval submodule is connected to the feature extraction submodule and is used to perform a neighbor retrieval of the query image using an image retrieval algorithm to obtain a neighbor retrieval pair of the query image; The feature matching submodule is connected to the image retrieval submodule and is used to establish a 2D-2D matching relationship between the query image and the database image; And the positioning submodule obtains the positioning result based on the positioning algorithm of the essential matrix.

[0015] Compared with the prior art, the present invention has the following beneficial effects: In this indoor visual positioning method and positioning system based on the occlusion removal algorithm, target detection, image restoration and feature matching technology are combined to improve the positioning accuracy and robustness.

[0016] By detecting occluders and removing them from all query images, we ensure that the pixels of the occluded objects will not interfere with subsequent processing, and provide a clear target area for the image restoration process. In this way, we can effectively isolate the occluded objects and provide strong support for subsequent image restoration algorithms.

[0017] When performing image restoration, the image restoration model MAT (Mask-Aware Transformer for Large Hole Image Inpainting) is used. This model combines the advantages of Transformer and convolution, and by introducing conditional long-distance interactions, it can effectively reconstruct image details in the presence of occlusion, thereby improving the accuracy and robustness of the visual positioning system.

[0018] Through the application of the MAT algorithm, the occluded image has been effectively repaired, which is not only more visually coherent, but also provides high-quality input data for subsequent visual positioning tasks. The MAT algorithm performs well in processing complex occlusion and high-resolution images, providing important support for the high accuracy and high robustness of the visual positioning system.

[0019] In this indoor visual positioning method based on occlusion removal algorithm, the SuperPoint feature extraction algorithm and the LightGlue feature matching algorithm are combined. This combination greatly improves the robustness and accuracy of feature detection and matching. After matching the feature points, the five-point method and RANSAC algorithm are used to estimate the essential matrix. Among them, the five-point method is an effective nonlinear method. The specific form of the essential matrix is ​​derived by solving the geometric constraints of five pairs of matching points. The RANSAC algorithm estimates the essential matrix by iteratively randomly selecting subsets from the feature matching set, and uses consistency checks to eliminate false matches, thereby improving the robustness and accuracy of the estimation. After solving the essential matrix, the relative rotation and translation are obtained by performing singular value decomposition on it, and the decomposed rotation matrix and translation vector will be further used to calculate the absolute pose of the query image. When calculating the absolute pose, the epipolar constraint is introduced to ensure the accuracy of the estimation result.

[0020] The use of the essential matrix can effectively fuse multi-view information and improve the accuracy of 3D reconstruction and positioning. Indoor environments usually have more occlusions and unstructured features. Single-view methods are difficult to provide sufficient constraints, while multi-view methods can provide more constraints through multi-angle view constraints, thereby significantly reducing positioning errors and improving positioning accuracy and robustness. In indoor visual positioning systems, this algorithm complements the image restoration method to jointly improve the positioning accuracy and robustness of the system, ensuring accurate indoor positioning in complex occlusion environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a flow chart of the indoor visual positioning method based on the occlusion removal algorithm.

[0022] Figure 2 This is the principle block diagram of the indoor visual positioning system based on the occlusion removal algorithm.

[0023] Figure 3 The following is a graph showing the cumulative distribution function (CDF) of the positioning error in the chess scene on the 7Scenes public dataset.

[0024] Figure 4 This is the cumulative distribution function (CDF) curve for the corridor 1 scenario.

[0025] Figure 5 This is the cumulative distribution function (CDF) curve for the corridor 2 scenario.

[0026] Figure 6 This is the cumulative distribution function (CDF) curve for the corridor 3 scenario. DETAILED DESCRIPTION

[0027] Figures 1 to 6 The best embodiment of the present invention is shown below in conjunction with the attached Figures 1 to 6 The present invention is further described.

[0028] like Figure 1 As shown, an indoor visual positioning method based on an occlusion removal algorithm includes the following steps: Step 1, database image feature extraction; The SuperPoint algorithm and NetVLAD algorithm are used to extract the feature information of the database image, and the feature information is saved for use in image matching and positioning.

[0029] First, reference points are set in the indoor environment and image data is collected at the reference points. Secondly, feature extraction is performed on the database image. Since the image retrieval module and the pose estimation module require different feature information, two different feature extraction algorithms are used for processing (SuperPoint and NetVLAD). Finally, the extracted feature information and the corresponding coordinate information are stored to form a feature file for subsequent image matching and positioning.

[0030] Step 2: occluded object detection; In order to study the impact of image occlusion on the visual positioning system, an occluded image dataset was constructed in this indoor visual positioning system based on the occlusion removal algorithm. First, 15 types of occluded objects, totaling 363 objects, were selected. They were obtained by cutout in the coco128 dataset. The cutout process involves separating the object of interest from the image to obtain an object image with a transparent background. Then, mask images of these objects are generated by retaining the pixels of the occluded objects and setting the pixel values ​​of the remaining positions to (255, 255, 255).

[0031] When generating the occluded image dataset, first load the images of the 7Scenes dataset; then load the images of all occluded objects and their corresponding masks; next, randomly select an occluded object and its mask for each query and database image, and add random types and degrees of blur information to it, including no blur, box blur, Gaussian blur, and motion blur. Then place the occluded object with blurred information at random positions on the image with random scales and rotation angles to generate the occluded image. Finally, the occluded images generated one-to-one for all images constitute the occluded image dataset.

[0032] The process of occluded object detection includes the following steps: Step 2-1, read the bounding box information in the detection result and extract the pixel area of ​​the occluded object.

[0033] The YOLOv5 model pre-trained on the COCO dataset is loaded and further fine-tuned on a specific occluded object dataset. The fine-tuned model is then used to detect the query image and generate bounding boxes and category information containing the occluded objects. After detecting the occluded object, the next step is to reset the pixels in the occluded object area for subsequent image restoration and visual positioning processing.

[0034] Step 2-2, reset the pixels in the area to pure black pixels (0, 0, 0) to indicate the position of the occluded object; at the same time, generate a mask image corresponding to the query image, in which the pixels in the occluded object area are set to pure black pixels, and the pixels in other areas are set to pure white pixels (255, 255, 255).

[0035] This step ensures that the pixels of the occluded object will not interfere with subsequent processing, and provides a clear target area for the image restoration process. In this way, the occluded object can be effectively isolated and provide strong support for the subsequent image restoration algorithm.

[0036] Step 3, image restoration; In order to solve the problem of large-area pixel loss and support diverse generation, the image restoration model MAT (Mask-Aware Transformer for Large Hole Image Inpainting) is used. This model combines the advantages of Transformer and convolution, and introduces conditional long-distance interaction to enable it to effectively reconstruct image details in the presence of occlusion, thereby improving the accuracy and robustness of the visual positioning system.

[0037] Through the application of the MAT algorithm, the occluded image has been effectively repaired, which is not only more visually coherent, but also provides high-quality input data for subsequent visual positioning tasks. The MAT algorithm performs well in processing complex occlusion and high-resolution images, providing important support for the high accuracy and high robustness of the visual positioning system.

[0038] Step 4: Indoor positioning; In the process of performing this step and performing indoor positioning, the following steps are included: Step 4-1, extract the feature information of the query image; perform feature extraction on the query image after occlusion removal, and use the SuperPoint algorithm and NetVLAD algorithm again to extract and save the feature information of the restored query image for use in image matching and positioning.

[0039] Step 4-2, performing image retrieval; performing a neighbor retrieval of the query image using an image retrieval algorithm based on the database image feature information and the query image feature information to obtain a neighbor retrieval pair of the query image; In the image retrieval process, the global feature vector of the image is generated with the help of algorithms such as CNN, and the similarity measurement function is used to calculate its similarity with each image in the database. Based on these similarities, the pairing relationship between all query images and database images is obtained.

[0040] The Patch-NetVLAD algorithm is used to measure the spatial and appearance consistency of image pairs by calculating the similarity scores between them, and finally obtain an image pair consisting of the query image and its first k neighbor images. This algorithm greatly reduces the computational cost caused by cross-matching of patch features while ensuring retrieval accuracy.

[0041] Step 4-3, perform feature matching; According to the feature information of database image, query image and nearest neighbor retrieval pairs, the LightGlue algorithm is used to perform feature matching and establish a 2D-2D matching relationship between the query image and the database image.

[0042] Step 4-4, perform indoor positioning; Using the indoor visual positioning algorithm based on the essential matrix, it uses the geometric relationship between multiple viewpoints in the image to achieve accurate 3D reconstruction and positioning. Based on the 2D-2D matching information, the five-point method in the RANSAC cycle is used to calculate the essential matrix of the image pair; then, through the epipolar constraint, the singular value decomposition (SVD) and RANSAC algorithm are used, combined with the known absolute pose and essential matrix of the database image, to finally determine the positioning result of the query camera.

[0043] In this indoor visual positioning method based on occlusion removal algorithm, the SuperPoint feature extraction algorithm and the LightGlue feature matching algorithm are combined. This combination greatly improves the robustness and accuracy of feature detection and matching. After matching the feature points, the five-point method and RANSAC algorithm are used to estimate the essential matrix. Among them, the five-point method is an effective nonlinear method. The specific form of the essential matrix is ​​derived by solving the geometric constraints of five pairs of matching points. The RANSAC algorithm estimates the essential matrix by iteratively randomly selecting subsets from the feature matching set, and uses consistency checks to eliminate false matches, thereby improving the robustness and accuracy of the estimation. After solving the essential matrix, the relative rotation and translation are obtained by performing singular value decomposition on it, and the decomposed rotation matrix and translation vector will be further used to calculate the absolute pose of the query image. When calculating the absolute pose, the epipolar constraint is introduced to ensure the accuracy of the estimation result.

[0044] The use of the essential matrix can effectively fuse multi-view information and improve the accuracy of 3D reconstruction and positioning. Indoor environments usually have more occlusions and unstructured features. Single-view methods are difficult to provide sufficient constraints, while multi-view methods can provide more constraints through multi-angle view constraints, thereby significantly reducing positioning errors and improving positioning accuracy and robustness. In indoor visual positioning systems, this algorithm complements the image restoration method to jointly improve the positioning accuracy and robustness of the system, ensuring accurate indoor positioning in complex occlusion environments.

[0045] like Figure 2 As shown, a positioning system for implementing the above-mentioned indoor visual positioning method based on occlusion removal algorithm includes an offline unit and an online unit. The offline unit includes an offline image acquisition module and an offline feature extraction module. The output of the offline image acquisition module is connected to the input of the offline feature extraction module. The above-mentioned step 1 is performed through the offline image acquisition module and the offline feature extraction module.

[0046] The online unit includes an online image acquisition module, an occlusion elimination module, an image restoration module and an indoor positioning module. The online image acquisition module is connected to the occlusion elimination module, the occlusion elimination module is connected to the image restoration module, the image restoration module is connected to the indoor positioning module, and the offline feature extraction module is also connected to the indoor positioning module.

[0047] Step 4-1 in the above step 4 is performed through the online image acquisition module. The occlusion removal module is provided with an occlusion object detection submodule and a pixel reset submodule, which are respectively used to implement the above steps 2-1 and 2-2. The image repair module is used to implement the above step 3, and the pixel reset submodule and the image repair module are simultaneously connected to the indoor positioning module.

[0048] In the indoor positioning module, including feature extraction submodule, image retrieval submodule, feature matching submodule and positioning submodule, the feature extraction submodule, image retrieval submodule, feature matching submodule and positioning submodule respectively execute the above steps 4-1 to step 4-4, and finally the positioning submodule executes the positioning algorithm based on the essential matrix to obtain the positioning result.

[0049] Through an example, the indoor visual positioning method and positioning system based on the occlusion removal algorithm are further verified: Using the 7Scenes_occ, 7Scenes_disocc and 7Scenes datasets, the system's positioning results include the median absolute position error (unit: m) and median absolute rotation error (unit: °) of each scene, as well as the average median positioning error of all scenes (including chess, fire, heads, office, pumpkin, red kitchen and stairs scenes). The error thresholds set by predecessors (0.25m, 2°), (0.5m, 5°), and (5m, 10°) are used, and the pass rate of the query image under each threshold is output.

[0050] Table 1 shows the ablation experiment of the occlusion removal module. The experiment is divided into three groups: the first group uses the 7Scenes_occ dataset, turns off the occlusion removal module, and directly performs positioning; the second group uses the 7Scenes_occ dataset, turns on the occlusion removal module, and performs positioning after occlusion removal processing; the third group uses the 7Scenes dataset. Although there are no occluded objects in the dataset, the occlusion removal module is still turned on to ensure the fairness of the experiment. In addition, the image retrieval algorithm used in the three groups of experiments is DenseVLAD. As shown in Table 1, the results of the second and third groups are similar, while the positioning accuracy of the first group is poor due to occlusion. Therefore, directly using occluded images will affect the positioning accuracy. After occlusion removal, the 7Scenes_occ dataset shows a positioning accuracy similar to that of the 7Scenes dataset, which proves the effectiveness of the occlusion removal module. Appendix Figure 3 The cumulative distribution function (CDF) curve of positioning error in the chess scene is shown, proving that the occlusion removal module significantly improves the positioning accuracy.

[0051] Table 1 Ablation experiment of occlusion removal module Note: In the first (), the first value is the median absolute position error, unit: m; the second value is the median absolute rotation error, unit: (°). In the second (), the three percentages are the pass rates of the query image under each threshold, unit: (%) In order to verify the improved indoor visual positioning algorithm proposed in this invention, a variety of advanced algorithms were experimentally compared using self-made LH data. The algorithms involved in the comparison are as follows: 1) A five-point algorithm based on SIFT (referred to as sift_5pt), which achieved excellent positioning results on the 7Scenes dataset; 2) An algorithm based on efficient image retrieval (referred to as retrieval), designed a fast retrieval algorithm, and combined with deep learning features to achieve good positioning results in 7Scenes; 3) The proposed algorithm based on SuperPoint and SuperGlue (referred to as sp+sg_5pt), which has more advantages in positioning accuracy and has become an advanced method in the current field of indoor positioning.

[0052] On this basis, this indoor visual positioning method based on the occlusion removal algorithm proposes an improved positioning algorithm based on SuperPoint and LightGlue (referred to as sp+lg_5pt). Figures 4 to 6 The cumulative distribution function (CDF) curves of these algorithms in three scenarios of LH dataset corridor 1 (corridor1), corridor 2 (corridor2) and corridor 3 (corridor3) are shown respectively. In addition, the quantitative evaluation results of the four algorithms are shown in Table 2. Figures 4 to 6 It can be clearly seen from the results in Table 2 that the positioning algorithm based on SuperPoint+LightGlue proposed in this indoor visual positioning method based on the occlusion removal algorithm shows higher positioning accuracy in all three scenes of the LH dataset.

[0053] Table 2 Comparison of positioning algorithm performance under LH dataset From the above, it can be seen that the indoor visual positioning system based on occlusion removal algorithm proposed by this indoor visual positioning method provides an effective solution to solve the impact of occluding objects on positioning accuracy. By combining target detection, image restoration and feature matching technology, the system not only improves the accuracy and robustness of positioning, but also provides new ideas for visual positioning problems in complex indoor environments.

[0054] The above is only a preferred embodiment of the present invention, and does not limit the present invention in other forms. Any technician familiar with the profession may use the above disclosed technical content to change or modify it into an equivalent embodiment with equivalent changes. However, any simple modification, equivalent change and modification made to the above embodiment according to the technical essence of the present invention without departing from the technical solution of the present invention still belongs to the protection scope of the technical solution of the present invention.

Claims

1. An indoor visual positioning method based on an occlusion removal algorithm, characterized in that: The steps include: Step 1, query the image in the database, extract the feature information of the database image, and save the feature information; Step 2: Detect occluders and remove all occluders in the query image. Step 3: Use the image restoration model to perform image restoration and restore the query image; Step 4: Perform indoor positioning.

2. The indoor visual positioning method based on the occlusion removal algorithm according to claim 1 is characterized in that: When executing step 2, it includes: constructing an occlusion image dataset; performing occlusion detection based on the occlusion image dataset; and resetting pixels to eliminate occlusions in the query image.

3. The indoor visual positioning method based on the occlusion removal algorithm according to claim 2 is characterized in that: When constructing the occluded image dataset: several objects are obtained in the coco128 dataset by matting. The matting process involves separating the object of interest from the image to obtain an object image with a transparent background; then, mask images of these objects are generated by retaining the pixels of the object and setting the pixel values ​​of the remaining positions to (255, 255, 255); Load the images of the 7Scenes dataset; then, load the images of all occluded objects and their corresponding masks; For each query and database image, an occluded object and its mask are randomly selected, and blur information of random type and degree is added to it. Then, the occluded object with blurred information is placed at a random position on the image with random scale and rotation angle to generate an occluded image. Finally, the occluded images generated one-to-one for all images constitute the occluded image dataset.

4. The indoor visual positioning method based on the occlusion removal algorithm according to claim 2 is characterized in that: The steps for occluded object detection based on the occluded image dataset are: first read the bounding box information in the detection result, extract the pixel area of ​​the occluded object, load the YOLOv5 model pre-trained on the COCO dataset, detect the query image, generate the bounding box and category information containing the occluded object, and complete the occluded object detection.

5. The indoor visual positioning method based on the occlusion removal algorithm according to claim 4 is characterized in that: The process of pixel resetting to eliminate occluders in the query image is as follows: the pixels in the occluded object area are reset to pure black pixels (0, 0, 0) to indicate the position of the occluded object; at the same time, a mask image corresponding to the query image is generated, in which the pixels in the occluded object area are set to pure black pixels, and the pixels in other areas are set to pure white pixels (255, 255, 255).

6. The indoor visual positioning method based on the occlusion removal algorithm according to claim 1, characterized in that: Step 4 includes the following steps: Step 4-1, extracting and saving feature information of the query image restored in step 3; Step 4-2, performing image retrieval; performing a neighbor retrieval of the query image using an image retrieval algorithm based on the database image feature information and the query image feature information to obtain a neighbor retrieval pair of the query image; Step 4-3, based on the database image feature information, the query image feature information and the nearest neighbor retrieval pair, use the LightGlue algorithm to perform feature matching and establish a 2D-2D matching relationship between the query image and the database image; Step 4-4, using the indoor visual positioning algorithm based on the essential matrix; based on the 2D-2D matching information, the five-point method in the RANSAC cycle is used to calculate the essential matrix of the image pair; then, through the epipolar constraint condition, the singular value decomposition and RANSAC algorithm are used, combined with the known absolute pose and essential matrix of the database image, to finally determine the positioning result.

7. The indoor visual positioning method based on the occlusion removal algorithm according to claim 1, characterized in that: When executing step 1, first set the reference point in the indoor environment and collect image data at the reference point; secondly, use the SuperPoint algorithm and NetVLAD algorithm to extract features from the database image; finally, store the extracted feature information and the corresponding coordinate information to form a feature file.

8. The indoor visual positioning method based on the occlusion removal algorithm according to claim 1, characterized in that: In step 3, the image restoration model MAT is used to perform image restoration.

9. A positioning system for implementing the indoor visual positioning method based on the occlusion removal algorithm according to any one of claims 1 to 8, characterized in that: It includes offline units and online units, where the offline units include: Offline image acquisition module, used to query images in the database; The offline feature extraction module is connected to the offline image acquisition module and is used to extract feature information of the database image and save the feature information; Online units include: An online image acquisition module, used to obtain query images; The occlusion removal module is connected to the online image acquisition module and is used to detect occlusions and remove occlusions in all query images; The indoor positioning module is connected to the occlusion removal module and the offline feature extraction module and is used for indoor positioning.

10. The positioning system according to claim 9, characterized in that: The indoor positioning module includes: A feature extraction submodule is used to obtain feature information of the restored query image; The image retrieval submodule is connected to the feature extraction submodule and is used to perform a neighbor retrieval of the query image using an image retrieval algorithm to obtain a neighbor retrieval pair of the query image; The feature matching submodule is connected to the image retrieval submodule and is used to establish a 2D-2D matching relationship between the query image and the database image; And the positioning submodule obtains the positioning result based on the positioning algorithm of the essential matrix.

Images