Method, device, equipment and storage medium for extracting candidate feature points of image
By parallelizing the differential images of the compressed and decompressed line laser scan images, the feature candidate points are extracted, and the problem of low extraction efficiency of feature candidate points at high frame rate in the prior art mid-line laser scanner is solved, and more efficient data transmission and Feature extraction.
Patent Information
- Application Number
- CN202210729356.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-06-24
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2042-06-24
AI Technical Summary
In the prior art, the linear laser scanner needs to wait for all images to be decompressed before extracting feature candidate points during high frame rate scanning, resulting in low extraction efficiency of feature candidate points.
The compressed first-frame image and multi-frame differential image sent by the receiving line laser scanning device are processed in parallel during the decompression process to extract feature candidate points, and use the differential image to divide the point set and feature point candidate areas to extract feature candidates to extract feature candidates. To avoid waiting for all differential images to be decompressed.
The extraction efficiency of feature candidate points is improved, efficient data transmission between the line laser scanning device and the processing device is ensured, and the extraction time of feature candidate points is reduced.
Smart Images

Figure CN114972133B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of image processing technology, and in particular to a method, device, equipment and storage medium for extracting candidate image feature points. Background Art
[0002] Line laser is a common 3D scanner light source. Line laser-based 3D measurement technology is a non-contact measurement method that can achieve large-area, high-resolution 3D object surface morphology measurement through high-speed laser scanning.
[0003] At present, when scanning at a high frame rate, in order to obtain efficient data transmission efficiency, the laser scanner needs to compress the scanned original image (wherein the original image is a line laser scanning image of the object to be measured), and send the compressed original image to a processing device, which first decompresses the compressed original image and then analyzes and processes the decompressed original image to obtain parameter information such as brightness or contrast of the original image, and uses this parameter information to extract feature candidate points of the original image, so as to calculate the three-dimensional contour information of the object to be measured based on the extracted feature candidate points.
[0004] However, when using the above-mentioned feature candidate point extraction method, the feature candidate points can only be extracted based on the decompressed original image after the compressed original image is completely decompressed, which easily leads to low extraction efficiency of the feature candidate points. Summary of the invention
[0005] The purpose of the present application is to provide a method, device, equipment and storage medium for extracting candidate feature points of an image, in order to improve the efficiency of extracting candidate feature points on a line laser scanning image of a measured object, in order to address the deficiencies in the above-mentioned prior art.
[0006] To achieve the above purpose, the technical solution adopted in the embodiment of the present application is as follows:
[0007] In a first aspect, an embodiment of the present application provides a method for extracting candidate image feature points, which is applied to a processing device, wherein the processing device is communicatively connected to a line laser scanning device, and the method includes:
[0008] Acquire a compressed multi-frame laser scanning image sent by the line laser scanning device, wherein the multi-frame laser scanning image includes: a compressed first frame image, and a compressed multi-frame differential image, wherein the differential image is obtained from two adjacent frames of original images scanned by the line laser scanning device;
[0009] Decompressing the compressed first frame image to obtain a first original image, taking the first original image as an initial first original image; and extracting a set of feature candidate points of the first original image;
[0010] Decompressing the compressed multi-frame differential images to obtain at least one first differential image;
[0011] Determine a second original image according to the first original image and the first differential image;
[0012] Dividing the first differential image into a plurality of point sets, each of the point sets includes points belonging to the same numerical interval;
[0013] Determine a feature point candidate region of the second original image according to a feature candidate point set of the first original image and a plurality of point sets of the first differential image, wherein the second original feature point candidate region includes a plurality of candidate feature points;
[0014] extracting a set of candidate feature points of the second original image from the second original image according to the candidate feature point region of the second original image;
[0015] Determine whether the feature candidate point sets of the multiple frames of original images have been completely extracted. If not, use the second original image as the new first original image, use the adjacent differential image after the first differential image as the new first differential image, and re-execute the above steps of determining the second original image based on the first original image and the first differential image until the feature candidate point sets of the multiple frames of original images have been completely extracted.
[0016] Optionally, the multiple point sets include: a first point set, a second point set, and a third point set, the values of the points in the first point set are all less than or equal to a first preset threshold, the values of the points in the second point set are greater than the first preset threshold and less than a second preset threshold, the values of the points in the third point set are greater than or equal to the second preset threshold, and the second preset threshold is greater than the first preset threshold.
[0017] Optionally, determining the feature point candidate region of the second original image according to the feature candidate point set of the first original image and the plurality of point sets of the first differential image includes:
[0018] Determine a difference between the first original feature candidate point set and the first point set to obtain a first intermediate set;
[0019] Determine the intersection of the first intermediate set and the second point set to obtain a second intermediate set;
[0020] A union of the second intermediate set and the third point set is determined to obtain a feature point candidate region of the second original image.
[0021] Optionally, extracting a set of candidate feature points of the second original image from the second original image according to the candidate feature point region of the second original image includes:
[0022] Determine a plurality of maximum points of the second original image according to the laser line scanning direction of the line laser scanning device, wherein each maximum point is a row-direction maximum point in each row or a column-direction maximum point in each column of the second original image, the row-direction maximum point is a feature point with the largest brightness in each row, and the column-direction maximum point is a feature point with the largest brightness in each column;
[0023] A feature candidate point set of the second original image is extracted based on the multiple maximum value points and each candidate feature point in the second original feature point candidate area.
[0024] Optionally, extracting a set of candidate feature points of the second original image according to the multiple maximum points and each candidate feature point in the feature point candidate region of the second original image includes:
[0025] Traversing each candidate feature point in the feature point candidate area of the second original image, comparing the currently traversed candidate feature point with the corresponding maximum value point, and if the value of the currently traversed candidate feature point is greater than the value of the corresponding maximum value point, taking the currently traversed candidate feature point as a point in the feature candidate point set of the second original image;
[0026] Among them, if the maximum point corresponding to the currently traversed candidate feature point is a row-wise maximum point, then the maximum point corresponding to the currently traversed candidate feature point is the row-wise maximum point of the row where the currently traversed candidate feature point is located; if the maximum point corresponding to the currently traversed candidate feature point is a column-wise maximum point, then the maximum point corresponding to the currently traversed candidate feature point is the column-wise maximum point of the column where the currently traversed candidate feature point is located.
[0027] Optionally, before extracting the set of candidate feature points of the second original image from the second original image according to the candidate feature point region of the second original image, the method further includes:
[0028] Performing dilation processing on the feature point candidate region of the first differential image to obtain a processed feature point candidate region;
[0029] The step of extracting a set of candidate feature points of the second original image from the second original image according to the candidate feature point region of the first differential image comprises:
[0030] A feature candidate point set of the second original image is extracted from the second original image according to the processed feature point candidate area.
[0031] Optionally, extracting a set of candidate feature points of the first original image includes:
[0032] Determine multiple maximum points of the first original image according to the laser line scanning direction of the line laser scanning device, wherein each maximum point is a row-direction maximum point in each row or a column-direction maximum point in each column of the first original image, the row-direction maximum point is a feature point with the maximum brightness value in each row, and the column-direction maximum point is a feature point with the maximum brightness value in each column;
[0033] A plurality of maximum value points of the first original image are combined into a feature candidate point set of the first original image.
[0034] In a second aspect, an embodiment of the present application further provides an image feature candidate point extraction device, which is applied to a processing device, wherein the processing device is communicatively connected to a line laser scanning device, and the device comprises:
[0035] An acquisition module, used for acquiring a compressed multi-frame laser scanning image sent by the line laser scanning device, wherein the multi-frame laser scanning image includes: a compressed first frame image, and a compressed multi-frame differential image, wherein the differential image is obtained from two adjacent frames of original images scanned by the line laser scanning device;
[0036] A decompression module, used for decompressing the compressed first frame image to obtain a first original image;
[0037] A processing module, configured to use the first original image as an initial first original image;
[0038] An extraction module, used to extract a set of feature candidate points of the first original image;
[0039] The decompression module is further used to decompress the compressed multi-frame differential image to obtain at least one first differential image;
[0040] The processing module is further used to determine a second original image according to the first original image and the first differential image;
[0041] A division module, used for dividing the first differential image into a plurality of point sets, each of the point sets including points belonging to the same numerical interval;
[0042] The processing module is further configured to determine a feature point candidate region of the second original image based on a feature point candidate set of the first original image and a plurality of point sets of the first differential image, wherein the second original feature point candidate region includes a plurality of candidate feature points;
[0043] The extraction module is further used to extract a set of candidate feature points of the second original image from the second original image according to the candidate feature point region of the second original image;
[0044] The processing module is also used to determine whether the feature candidate point sets of the multiple frames of original images have been completely extracted. If not, the second original image is used as the new first original image, the adjacent differential image after the first differential image is used as the new first differential image, and the above-mentioned steps of determining the second original image based on the first original image and the first differential image are re-executed until the feature candidate point sets of the multiple frames of original images are completely extracted.
[0045] Optionally, the multiple point sets include: a first point set, a second point set, and a third point set, the values of the points in the first point set are all less than or equal to a first preset threshold, the values of the points in the second point set are greater than the first preset threshold and less than a second preset threshold, the values of the points in the third point set are greater than or equal to the second preset threshold, and the second preset threshold is greater than the first preset threshold.
[0046] Optionally, the processing module is further used to:
[0047] Determine a difference between the first original feature candidate point set and the first point set to obtain a first intermediate set;
[0048] Determine the intersection of the first intermediate set and the second point set to obtain a second intermediate set;
[0049] A union of the second intermediate set and the third point set is determined to obtain a feature point candidate region of the second original image.
[0050] Optionally, the extraction module is further used to:
[0051] Determine a plurality of maximum points of the second original image according to the laser line scanning direction of the line laser scanning device, wherein each maximum point is a row-direction maximum point in each row or a column-direction maximum point in each column of the second original image, the row-direction maximum point is a feature point with the largest brightness in each row, and the column-direction maximum point is a feature point with the largest brightness in each column;
[0052] A feature candidate point set of the second original image is extracted based on the multiple maximum value points and each candidate feature point in the second original feature point candidate area.
[0053] Optionally, the extraction module is further used to:
[0054] Traversing each candidate feature point in the feature point candidate area of the second original image, comparing the currently traversed candidate feature point with the corresponding maximum value point, and if the value of the currently traversed candidate feature point is greater than the value of the corresponding maximum value point, taking the currently traversed candidate feature point as a point in the feature candidate point set of the second original image;
[0055] Among them, if the maximum point corresponding to the currently traversed candidate feature point is a row-wise maximum point, then the maximum point corresponding to the currently traversed candidate feature point is the row-wise maximum point of the row where the currently traversed candidate feature point is located; if the maximum point corresponding to the currently traversed candidate feature point is a column-wise maximum point, then the maximum point corresponding to the currently traversed candidate feature point is the column-wise maximum point of the column where the currently traversed candidate feature point is located.
[0056] Optionally, the processing module is further used to:
[0057] Performing dilation processing on the feature point candidate region of the first differential image to obtain a processed feature point candidate region;
[0058] The extraction module is further used for:
[0059] A feature candidate point set of the second original image is extracted from the second original image according to the processed feature point candidate area.
[0060] Optionally, the extraction module is further used to:
[0061] Determine multiple maximum points of the first original image according to the laser line scanning direction of the line laser scanning device, wherein each maximum point is a row-direction maximum point in each row or a column-direction maximum point in each column of the first original image, the row-direction maximum point is a feature point with the maximum brightness value in each row, and the column-direction maximum point is a feature point with the maximum brightness value in each column;
[0062] A plurality of maximum value points of the first original image are combined into a feature candidate point set of the first original image.
[0063] In a third aspect, an embodiment of the present application further provides a processing device, comprising: a processor, a storage medium and a bus, wherein the storage medium stores machine-readable instructions executable by the processor. When the processing device is running, the processor and the storage medium communicate through the bus, and the processor executes the machine-readable instructions to perform the steps of the method provided in the first aspect.
[0064] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the method provided in the first aspect are executed.
[0065] The beneficial effects of this application are:
[0066] The embodiment of the present application provides an image feature candidate point extraction method, device, equipment and storage medium. In order to ensure efficient data transmission efficiency between a line laser scanning device and a processing device, the processing device directly receives the compressed first frame image and the compressed multi-frame differential image sent by the line laser scanning device. The processing device decompresses the compressed first frame image to obtain a first original image, and extracts a feature candidate point set of the first original image. In addition, the processing device first decompresses part of the compressed multi-frame differential image to obtain at least one first differential image, and then extracts a feature candidate point set of the second original image based on the first differential image that is closer to the front after decompression and the feature candidate point set of the first original image. At the same time, in the process of extracting the feature candidate point set of the second original image, the compressed other multi-frame differential images continue to be decompressed. This parallel processing method of image decompression and feature candidate point extraction does not require waiting for the compressed multi-frame differential images to be completely decompressed before starting to extract the feature candidate point set on each frame of the original image based on the decompressed multi-frame differential images, which greatly improves the extraction efficiency of the feature candidate points. BRIEF DESCRIPTION OF THE DRAWINGS
[0067] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.
[0068] Figure 1 A schematic diagram of the structure of a line laser scanning system provided in an embodiment of the present application;
[0069] Figure 2 A schematic diagram of the structure of a processing device provided in an embodiment of the present application;
[0070] Figure 3 A schematic diagram of a flow chart of a method for extracting candidate image feature points provided in an embodiment of the present application;
[0071] Figure 4 A schematic diagram of a flow chart of another method for extracting candidate image feature points provided in an embodiment of the present application;
[0072] Figure 5 A schematic diagram of a flow chart of another method for extracting candidate image feature points provided in an embodiment of the present application;
[0073] Figure 6 A schematic diagram of a flow chart of another method for extracting candidate image feature points provided in an embodiment of the present application;
[0074] Figure 7 A schematic diagram of a flow chart of another method for extracting candidate image feature points provided in an embodiment of the present application;
[0075] Figure 8 A schematic diagram of the structure of an image feature candidate point extraction device provided in an embodiment of the present application.
[0076] Icon: 100 - line laser scanning system; 101 - line laser scanning equipment; 102 - processing equipment. DETAILED DESCRIPTION
[0077] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It should be understood that the drawings in the present application only serve the purpose of explanation and description and are not used to limit the scope of protection of the present application. In addition, it should be understood that the schematic drawings are not drawn in real proportion. The flowchart used in this application shows the operations implemented according to some embodiments of the present application. It should be understood that the operations of the flowchart can be implemented out of sequence, and the steps without logical context can be reversed in order or implemented simultaneously. In addition, those skilled in the art can add one or more other operations to the flowchart under the guidance of the content of the present application, or remove one or more operations from the flowchart.
[0078] In addition, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.
[0079] It should be noted that the term "comprising" will be used in the embodiments of the present application to indicate the existence of the features declared thereafter, but does not exclude the addition of other features.
[0080] First of all, before elaborating on the technical solution provided by the present application, a brief introduction is first given to the architecture and working principle of the line laser scanning system involved in the present application.
[0081] refer to Figure 1 As shown, the line laser scanning system 100 includes: a line laser scanning device 101 and a processing device 102, and the line laser scanning device 101 is communicatively connected with the processing device 102. Among them, the line laser scanning device 101 may include: a line laser and an image acquisition device, and the processing device 102 may be a processing device such as a computer or a server, so as to implement the image feature candidate point extraction method provided in this application.
[0082] The working principle of the line laser scanning system 100 is: a line laser is emitted to the surface of the object to be measured by a line laser, and a line laser scanning image of the object to be measured is collected in real time by an image acquisition device, and the collected line laser scanning image is sent to the processing device 102. The processing device 102 analyzes and processes the received line laser scanning image to obtain the real three-dimensional contour information of the object to be measured.
[0083] Understandably, Figure 1 The structure shown is for illustration only. The line laser scanning system 100 may also include Figure 1 More or fewer components as shown, or with Figure 1 Different configurations are shown. Figure 1 Each component shown in the figure can be implemented by hardware, software or a combination thereof.
[0084] The following is a brief description of the structural schematic diagram of a processing device for executing the image feature candidate point extraction method provided in the present application.
[0085] like Figure 2 As shown, the processing device 102 includes a memory 201 and a processor 202. The memory 201 and the processor 202 are electrically connected to each other directly or indirectly to achieve data transmission or interaction. For example, these components can be electrically connected to each other through one or more communication buses or signal lines.
[0086] The memory 201 stores software function modules in the form of software or firmware. The processor 202 executes various functional applications and data processing by running the software programs and modules stored in the memory 201, thereby realizing the image feature candidate point extraction method in the embodiment of the present application.
[0087] The memory 201 may be, but is not limited to, a random access memory (RAM), a read only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), etc. The memory 201 is used to store a program, and the processor 202 executes the program after receiving an execution instruction.
[0088] The processor 202 may be an integrated circuit chip with the ability to process signals. The processor 202 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), and the like.
[0089] The implementation principle and corresponding beneficial effects of the image feature candidate point extraction method steps provided in the present application will be explained through a number of specific embodiments as follows.
[0090] Figure 3 A schematic diagram of a flow chart of a method for extracting candidate image feature points provided in an embodiment of the present application; optionally, the execution subject of the method may be as shown in Figure 1 processing equipment in .
[0091] It should be understood that in other embodiments, the order of some steps in the method for extracting candidate image feature points can be interchanged according to actual needs, or some steps can be omitted or deleted. Figure 3 As shown, the method includes:
[0092] S301, obtaining a compressed multi-frame laser scanning image sent by a line laser scanning device.
[0093] The multiple frames of laser scanning images include: a compressed first frame image and compressed multiple frames of differential images, and the differential image is obtained from two adjacent frames of original images scanned by a line laser scanning device.
[0094] In this embodiment, for example, after the line laser scanning device acquires multiple frames of original images of the object to be measured, wherein the original images are line laser scanning images of the object to be measured, it is also necessary to obtain a differential image between two adjacent frames of original images according to two adjacent frames of original images in the multiple frames of original images, that is, to subtract the previous frame of original image from the current frame of original image. For example, the line laser scanning device acquires n frames of original images, and obtains a first differential image D1 between the first frame of original image F1 and the second frame of original image F2 according to the pixel values of each pixel point in the first frame of original image F1 and the pixel values of each pixel point in the second frame of original image F2, that is, D1 = F2-F1; similarly, a second differential image D2 between the second frame of original image F2 and the third frame of original image F3 can also be obtained, that is, according to the adjacent two frames of original images of the n frames of original images, an n-1 frame of differential image can be obtained.
[0095] At the same time, the line laser scanning device also needs to compress the first frame original image (i.e. the first frame original image F1) and the n-1 frames of differential images obtained above to obtain the compressed first frame image and the compressed multi-frame differential images, and send the compressed first frame image and the compressed multi-frame differential images to the processing device, so that the processing device can extract the feature candidate point set on each frame of the original image based on the compressed first frame image and the compressed multi-frame differential images.
[0096] In this embodiment, in order to ensure efficient data transmission efficiency between the line laser scanning device and the processing device, the line laser scanning device can send the compressed first frame image and the compressed multiple frames of differential images to the processing device. In this way, the amount of data transmission between the line laser scanning device and the processing device can be greatly reduced.
[0097] It is worth noting that while the online laser scanning device collects the original image of the object to be measured, the processing device extracts the feature candidate points of a certain frame of the original image based on the currently acquired multiple frames of laser scanning images, that is, the image acquisition and feature candidate point extraction are processed in parallel. For example, the line laser scanning device collects the first frame of the original image of the object to be measured, compresses the first frame of the original image to obtain the compressed first frame image, and transmits the compressed first frame image to the processing device. While the processing device decompresses the compressed first frame image and extracts the feature candidate points, the line laser scanning device continues to collect the second frame of the original image of the object to be measured, and obtains the difference image of the second frame of the original image and the first frame of the original image based on the first frame of the original image and the second frame of the original image, that is, the first difference image, and compresses the first difference image, and then transmits the compressed first difference image to the processing device, and the processing device extracts the feature candidate points of the second frame of the original image based on the compressed first frame image and the compressed first difference image, and so on. This parallel processing method of collecting images and extracting feature candidate points can greatly improve the extraction efficiency of feature candidate points.
[0098] S302 , decompressing the compressed first frame image to obtain a first original image, taking the first original image as an initial first original image, and extracting a set of feature candidate points of the first original image.
[0099] The features of the first original image may include: laser lines and marking points, that is, a set of candidate laser line points or a set of candidate marking points of the first original image is extracted.
[0100] In this embodiment, the processing device first decompresses the compressed first frame image to obtain a first original image, that is, a first frame original image, and uses the first original image as an initial first original image.
[0101] Optionally, for the first frame of the original image, parameter information such as brightness and contrast of the first original image can be calculated, and these parameter information can be used to extract a set of feature candidate points of the first original image, and the set of feature candidate points of the first original image is recorded as L1.
[0102] S303: Decompress the compressed differential image frames to obtain at least one first differential image.
[0103] Optionally, the compressed n-1 frames of differential images may be decompressed at the same time, or the compressed n-1 frames of differential images may be decompressed in batches to obtain at least one first differential image.
[0104] In this embodiment, for example, the processing device can decompress the compressed n-1 frames of differential images in batches, so that the feature candidate point set of the second original image can be extracted based on the decompressed first differential image and the first original image, and then the feature candidate point set of the third original image can be extracted based on the decompressed second differential image and the second original image, etc., that is, the feature candidate point set of the earlier original image in the multiple frames of original images is extracted based on the earlier differential image after decompression and the first original image. In this way, it is not necessary to wait until all the compressed multiple frames of differential images are decompressed before starting to extract the feature candidate points on each frame of the original image in turn based on the decompressed multiple frames of differential images. The parallel processing method of image decompression and feature candidate point extraction is adopted, which greatly improves the extraction efficiency of the feature candidate points on each frame of the original image.
[0105] S304: Determine a second original image according to the first original image and the first differential image.
[0106] In this embodiment, for example, in the first round of cyclic processing, the first original image can be recorded as F1, and the first differential image can be recorded as D1, wherein the first differential image D1 is calculated by the line laser scanning device according to the first original image F1 and the second original image F2, and correspondingly, the first original image F1 and the first differential image D1 can also be used on the processing device to perform reverse reasoning calculation to obtain the second original image F2. For example, if the first differential image D1 = the second original image F2 - the first original image F1, then the second original image F2 = the first differential image D1 + the first original image F1.
[0107] S305: Divide the first differential image into multiple point sets.
[0108] Each point set includes points belonging to the same numerical interval.
[0109] The differential image is the difference between two adjacent image frames, that is, the change of features between two adjacent image frames can be determined according to the change of pixel values of each pixel point on the differential image.
[0110] In this embodiment, the pixel points on the first differential image that belong to the same pixel value interval can be divided into a point set according to the pixel value of each pixel point on the first differential image, that is, the first differential image can be divided into multiple point sets.
[0111] S306 : Determine a feature point candidate region of the second original image according to the feature candidate point set of the first original image and the plurality of point sets of the first differential image.
[0112] The feature point candidate region of the second original image includes a plurality of candidate feature points, the feature point candidate region contained in the second original image may be one or more, and each feature point candidate region may include one candidate feature point or a plurality of candidate feature points.
[0113] In this embodiment, the first differential image is determined by the first original image and the second original image, that is, the pixel value of each pixel in the first differential image can represent the change in the pixel value of the same pixel in the first original image and the second original image. Therefore, the feature point candidate area of the second original image can be determined by using the first differential image and the feature candidate point set of the first original image.
[0114] S307 . Extracting a set of candidate feature points of the second original image from the second original image according to the candidate feature point region of the second original image.
[0115] In this embodiment, the feature candidate points of the second original image can be directly extracted from the feature point candidate region of the second original image, and the extracted feature candidate points are combined into a feature candidate point set of the second original image. In this way, it is only necessary to extract the feature candidate points of the second original image from the feature point candidate region of the second original image, avoiding the tedious process of directly comparing and analyzing all the pixels on the entire second original image one by one to extract the feature candidate points of the second original image, greatly reducing the search area of the feature candidate points of the second original image, thereby accelerating the extraction efficiency of the feature candidate points of the second original image.
[0116] S308, judging whether the extraction of the feature candidate point sets of the multiple frames of original images is complete, if not, jumping to step S309, if yes, jumping to step S310.
[0117] S309: Use the second original image as a new first original image, use the adjacent differential image after the first differential image as a new first differential image, and re-execute steps S304-S310.
[0118] S310: If all feature candidate point sets of multiple frames of original images have been extracted, the process ends.
[0119] In this embodiment, after extracting the feature candidate point set of the second original image, the next round of loop processing can be entered, that is, in the next round of loop processing, the second original image is used as the new first original image, the adjacent differential image after the first differential image is used as the new first differential image, and steps S304-S310 are re-executed until the feature candidate point sets of multiple frames of original images are all extracted, and then the process ends.
[0120] In summary, an embodiment of the present application provides an image feature candidate point extraction method, which is applied to a processing device, and the method includes: obtaining a compressed multi-frame laser scanning image sent by a line laser scanning device, wherein the multi-frame laser scanning image includes: a compressed first frame image, and a compressed multi-frame differential image, and the differential image is obtained by scanning two adjacent frames of original images by the line laser scanning device; decompressing the compressed first frame image to obtain a first original image, using the first original image as an initial first original image, and extracting a feature candidate point set of the first original image; decompressing the compressed multi-frame differential image to obtain at least one first differential image; determining a second original image according to the first original image and the first differential image; dividing the first differential image into multiple point sets, Each point set includes points belonging to the same numerical interval; based on the feature candidate point set of the first original image and the multiple point sets of the first differential image, the feature point candidate area of the second original image is determined, wherein the feature point candidate area of the second original image includes multiple candidate feature points; based on the feature point candidate area of the second original image, the feature candidate point set of the second original image is extracted from the second original image; it is determined whether the feature candidate point sets of multiple frames of original images have been completely extracted, if not, the second original image is used as a new first original image, the adjacent differential image after the first differential image is used as a new first differential image, and the above steps after determining the second original image based on the first original image and the first differential image are re-executed until the feature candidate point sets of multiple frames of original images are completely extracted. In this scheme, in order to ensure efficient data transmission efficiency between the line laser scanning device and the processing device, the processing device directly receives the compressed first frame image and the compressed multi-frame differential images sent by the line laser scanning device, and the processing device decompresses the compressed first frame image to obtain the first original image, and extracts the feature candidate point set of the first original image. In addition, the processing device can first decompress the compressed part of the multi-frame differential images and obtain at least one first differential image, and then extract the feature candidate point set of the second original image based on the decompressed first differential image and the feature candidate point set of the first original image. At the same time, in the process of extracting the feature candidate point set of the second original image, the other compressed multi-frame differential images continue to be decompressed. This parallel processing method of image decompression and feature candidate point extraction does not need to wait for the compressed multi-frame differential images to be fully decompressed before starting to extract the feature candidate point set on each frame of the original image based on the decompressed multi-frame differential images, which greatly improves the extraction efficiency of the feature candidate points on each frame of the original image.
[0121] The following embodiment will be used to explain in detail how to determine the feature point candidate region of the second original image based on the feature candidate point set of the first original image and the multiple point sets of the first differential image in the above step S306.
[0122] First, a plurality of point sets of the first differential image are introduced.
[0123] Optionally, the plurality of point sets of the first differential image may include: a first point set B N 、The second point set C N 、The third point set D N , where the first point set B N The values of the points in the second point set C are all less than or equal to the first preset threshold t1. N The value of the point in is greater than the first preset threshold t1 and less than the second preset threshold t2. The third point set D N The value of the point in is greater than or equal to the second preset threshold t2, and the second preset threshold t2 is greater than the first preset threshold t1, that is, t2>t1.
[0124] In this embodiment, according to the first preset threshold t1 (t1<=0) and the second preset threshold t2 (t2>=0), the pixel values of each pixel on the first differential image are divided into three point sets, namely, the first point set B N 、The second point set C N 、The third point set D N , where the first point set B N represents the point set whose pixel values on the first differential image are less than the first preset threshold value t1, the second point set CN represents the point set whose pixel values on the first differential image are between the first preset threshold value t1 and the second preset threshold value t2, and the third point set D N The point set represents the pixel values of the pixels on the first differential image that are greater than the second preset threshold t2.
[0125] It should be noted that the first point set B N The pixel value contained in is less than the first preset threshold t1, that is, less than 0, that is, the first point set B N The pixel value included in represents that at the same pixel point, the pixel value of a feature in the second original image is smaller than the pixel value of the first original image. For example, a certain marking point appears at pixel point A in the first original image, but does not appear at pixel point A in the second original image (or the pixel point position of the marking point in the second original image changes), that is, the marking point disappears directly in the second original image at the same pixel point.
[0126] The second point set C NThe pixel values contained in the second point set C are between the first preset threshold value t1 and the second preset threshold value t2. N The pixel values included in represent that at the same pixel point, the pixel value of a feature in the second original image and the pixel value of the first original image do not vary much. For example, the pixel value of a laser line at pixel point A in the first original image is 120, and the pixel value of the laser line at pixel point A in the second original image is 121, that is, at the same pixel point, the pixel value of the laser line in the first original image and the second original image does not vary significantly.
[0127] The third point set D N The pixel value contained in is greater than or equal to the second preset threshold t2, that is, the third point set D N The pixel value included in represents that at the same pixel point, the pixel value of a feature in the second original image is greater than the pixel value of the first original image. For example, the pixel value of a marker point at pixel point A in the first original image is 0, and the pixel value of the marker point at pixel point A in the second original image is 50, that is, the marker point does not appear in the first original image, but appears in the second original image.
[0128] Optionally, refer to Figure 4 As shown, the above step S306 includes:
[0129] S401. Determine a difference between a feature candidate point set of a first original image and a first point set to obtain a first intermediate set.
[0130] S402: Determine the intersection of the first intermediate set and the second point set to obtain a second intermediate set.
[0131] S403: Determine the union of the second intermediate set and the third point set to obtain a feature point candidate region of the second original image.
[0132] In this embodiment, the first original feature candidate point set is denoted as L N-1 , then the first original feature candidate point set L N-1 With the first point set B N The difference set is L N-1 -B N , that is, directly take the first point set B N From the first original feature candidate point set L N-1 The first intermediate set is filtered out to obtain the first intermediate set. The first intermediate set represents filtering out the feature points that directly disappear in the second original image at the same pixel point.
[0133] Calculate the first intermediate set (L N-1 -B N ) and the second point set C NThe intersection of the two sets gives the second intermediate set (L N-1 -B N )∩C N , that is, to find the second point set C N The feature points that are the same as those in the first intermediate set are obtained to obtain the second intermediate set, thereby avoiding the duplication of feature candidate points.
[0134] Calculate the second intermediate set (L N-1 -B N )∩C N With the third point set D N The union of the feature point candidate region E of the second original image is obtained. N , that is, directly set the third point set D N The feature candidate point set added to the second intermediate set (L N-1 -B N ), and obtain the feature point candidate region E of the second original image N , which can also be written as E N =(L N-1 -B N )∩C N ∪D N , where L N-1 It is the set of candidate feature points on the N-1 frames of original image.
[0135] In this embodiment, when obtaining the feature point candidate area of the second original image, it is not necessary to calculate the contrast or brightness parameter information of the second original image. It is only necessary to compare and perform set operations on multiple point sets of the first differential image and the feature candidate point set of the first original image to obtain the feature point candidate area of the second original image, thereby reducing the amount of calculation so as to obtain the feature candidate point set more quickly.
[0136] The following embodiment will be used to explain in detail how to extract a set of candidate feature points of the second original image from the second original image according to the candidate feature point region of the second original image in the above S307.
[0137] Optionally, refer to Figure 5 As shown, the above step S307 includes:
[0138] S501. Determine a plurality of maximum value points of a second original image according to a laser line scanning direction of a line laser scanning device.
[0139] Among them, each maximum point is a row maximum point in each row or a column maximum point in each column of the second original image, the row maximum point is the feature point with the maximum brightness in each row, and the column maximum point is the feature point with the maximum brightness in each column.
[0140] In one achievable manner, for example, when the laser line scanning direction of the line laser scanning device is along the row direction, the column-wise maximum points in the second original image are used to determine a plurality of maximum points in the second original image.
[0141] In another achievable manner, for example, when the laser line scanning direction of the line laser scanning device is along the column direction, the row-wise maximum points in the second original image are used to determine a plurality of maximum points in the second original image.
[0142] In this embodiment, when the laser line scanning direction of the line laser scanning device is along the row direction, the brightness parameters in each column of the second original image are calculated, and the maximum value of the brightness parameters in each column is used as the column-wise maximum point of the column; or, when the laser line scanning direction of the line laser scanning device is along the column direction, the brightness parameters in each row of the second original image are calculated, and the maximum value of the brightness parameters in each row is used as the row-wise maximum point of the row.
[0143] S502: Extract a feature candidate point set of the second original image according to the multiple maximum points and each candidate feature point in the second original feature point candidate region.
[0144] In this embodiment, when the laser line scanning direction of the line laser scanning device is along the row direction, the column-wise maximum points in the second original image are used to analyze and compare all the candidate feature points on each column in the second original feature point candidate area, and multiple feature candidate points of the second original image are extracted, and a feature candidate point set of the second original image is obtained; or, when the laser line scanning direction of the line laser scanning device is along the column direction, the row-wise maximum points in the second original image are used to analyze and compare all the candidate feature points on each row in the second original feature point candidate area, and multiple feature candidate points of the second original image are extracted, and a feature candidate point set of the second original image is obtained.
[0145] In this embodiment, when obtaining the set of feature candidate points of the second original image, it is not necessary to calculate the contrast parameter information of the second original image. It is only necessary to screen the candidate feature points in the second original feature point candidate area according to the row maximum points in each row or the column maximum points in each column of the second original image. The feature candidate points of the second original image can be extracted more quickly, avoiding the screening process of all feature candidate points in the entire second original image, reducing the amount of calculation, and thus improving the extraction efficiency of the feature candidate points.
[0146] Optionally, in the above step S502, extracting a set of candidate feature points of the second original image according to the plurality of maximum points and each candidate feature point in the candidate feature point region of the second original image includes:
[0147] Traverse each candidate feature point in the feature point candidate area of the second original image, compare the currently traversed candidate feature point with the corresponding maximum point, if the value of the currently traversed candidate feature point is greater than the value of the corresponding maximum point, then take the currently traversed candidate feature point as a point in the feature candidate point set of the second original image.
[0148] Among them, if the maximum point corresponding to the currently traversed candidate feature point is a row-wise maximum point, then the maximum point corresponding to the currently traversed candidate feature point is the row-wise maximum point of the row where the currently traversed candidate feature point is located; if the maximum point corresponding to the currently traversed candidate feature point is a column-wise maximum point, then the maximum point corresponding to the currently traversed candidate feature point is the column-wise maximum point of the column where the currently traversed candidate feature point is located.
[0149] In this embodiment, when the laser line scanning direction of the line laser scanning device is along the row direction, the candidate feature points in all columns in the feature point candidate area of the second original image are traversed, and the currently traversed candidate feature point is compared with the column-wise maximum point of the column in which it is located. If the pixel value of the currently traversed candidate feature point is greater than the column-wise maximum point of the column in which it is located, the currently traversed candidate feature point is used as a point in the feature candidate point set of the second original image.
[0150] Similarly, when the laser line scanning direction of the line laser scanning device is along the column direction, the candidate feature points of all rows in the feature point candidate area of the second original image are traversed, and the candidate feature point currently traversed is compared with the row-wise maximum point of the row in which it is located. If the pixel value of the candidate feature point currently traversed is greater than the column-wise maximum point of the row in which it is located, the candidate feature point currently traversed is used as a point in the feature candidate point set of the second original image. In this way, all candidate features in the feature point candidate area of the second original image can be traversed and analyzed to extract the feature candidate point set of the second original image, thereby avoiding the loss of feature points in the process of extracting the feature candidate points of the second original image.
[0151] Optionally, refer to Figure 6 As shown, before executing the above S307 to extract the feature candidate point set of the second original image from the second original image according to the feature point candidate area of the second original image, it also includes:
[0152] S601 , dilating the feature point candidate region of the second original image to obtain a processed feature point candidate region.
[0153] The above step S307 extracts a set of candidate feature points of the second original image from the second original image according to the candidate feature point region of the second original image, including:
[0154] S602: Extracting a feature candidate point set of the second original image from the second original image according to the processed feature point candidate region.
[0155] In this embodiment, in order to avoid omission of feature candidate points in each frame of the original image, it is proposed that the feature point candidate area E of the second original image can also be N Perform an expansion operation to convert the feature point candidate area E of the second original image N The neighboring points of the inlier point are also used as candidate points, that is, the processed feature point candidate region is obtained; then, based on the processed feature point candidate region, the feature candidate point set L of the second original image is extracted from the second original image. N .
[0156] Optionally, refer to Figure 7 As shown, the step S302 extracts a set of candidate feature points of the first original image, including:
[0157] S701. Determine a plurality of maximum points of a first original image according to a laser line scanning direction of a line laser scanning device.
[0158] Among them, each maximum point is a row maximum point in each row or a column maximum point in each column of the first original image, the row maximum point is a feature point with the largest brightness value in each row, and the column maximum point is a feature point with the largest brightness value in each column.
[0159] S702: Combine multiple maximum value points of the first original image into a feature candidate point set of the first original image.
[0160] It has been mentioned in the above embodiments that when the laser line scanning direction of the line laser scanning device is along the row direction, the column-wise maximum points in the second original image can be used to determine multiple maximum points in the second original image; or, when the laser line scanning direction of the line laser scanning device is along the column direction, the row-wise maximum points in the second original image can be used to determine multiple maximum points in the second original image.
[0161] In this embodiment, when the laser line scanning direction of the line laser scanning device is along the column direction, the row-wise maximum point in the first original image is used to determine multiple maximum points in the first original image. Specifically, by calculating the brightness values of all pixels in each row in the first original image, the pixel with the largest brightness value is used as the row-wise maximum point of the row where the pixel is located, and the row-wise maximum points of each row in the first original image are used as multiple maximum points in the first original image, that is, the multiple maximum points of the first original image are combined into a feature candidate point set of the first original image.
[0162] Similarly, when the laser line scanning direction of the line laser scanning device is along the row direction, the column-wise maximum points in the first original image can be used to determine multiple maximum points in the first original image. Specifically, the brightness values of all pixels in each column in the first original image can be calculated, and the pixel with the largest brightness value can be used as the column-wise maximum point of the column where the pixel is located, and the column-wise maximum points of each column in the first original image can be used as multiple maximum points in the first original image, that is, the multiple maximum points of the first original image can be combined into a feature candidate point set of the first original image.
[0163] In this embodiment, when obtaining the set of feature candidate points of the first original image, it is not necessary to calculate the contrast parameter information of the first original image. It is only necessary to extract the feature candidate points of the first original image more quickly based on the row-wise maximum points in each row or the column-wise maximum points in each column of the first original image, thereby reducing the amount of calculation and improving the efficiency of extracting the feature candidate points.
[0164] Based on the same inventive concept, an image feature candidate point extraction device corresponding to the image feature candidate point extraction method is also provided in the embodiment of the present application. Since the principle of solving the problem by the device in the embodiment of the present application is similar to the above-mentioned image feature candidate point extraction method in the embodiment of the present application, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be repeated.
[0165] refer to Figure 8 As shown, the embodiment of the present application also provides an image feature candidate point extraction device, which is applied to a processing device, and the processing device is communicatively connected with a line laser scanning device, and the device includes:
[0166] The acquisition module 801 is used to acquire the compressed multi-frame laser scanning images sent by the line laser scanning device, wherein the multi-frame laser scanning images include: a compressed first frame image and a compressed multi-frame differential image, where the differential image is obtained from two adjacent frames of original images scanned by the line laser scanning device;
[0167] The decompression module 802 is used to decompress the compressed first frame image to obtain a first original image;
[0168] A processing module 803 is used to use the first original image as an initial first original image;
[0169] An extraction module 804 is used to extract a set of feature candidate points of the first original image;
[0170] The decompression module 802 is further used to decompress the compressed multi-frame differential image to obtain at least one first differential image;
[0171] The processing module 803 is further used to generate a second original image based on the first original image and the first differential image;
[0172] A division module 805, configured to divide the first differential image into a plurality of point sets, each point set including points belonging to the same numerical interval;
[0173] The processing module 803 is further used to determine a feature point candidate region of the second original image according to the feature candidate point set of the first original image and the multiple point sets of the first differential image, wherein the second original feature point candidate region includes multiple candidate feature points;
[0174] The extraction module 804 is further used to extract a set of candidate feature points of the second original image from the second original image according to the candidate feature point region of the second original image;
[0175] The processing module 803 is also used to determine whether the feature candidate point sets of the multiple frames of original images have been completely extracted. If not, the second original image is used as the new first original image, the adjacent differential image after the first differential image is used as the new first differential image, and the above steps of determining the second original image based on the first original image and the first differential image are re-executed until the feature candidate point sets of the multiple frames of original images are completely extracted.
[0176] Optionally, the multiple point sets include: a first point set, a second point set, and a third point set, the values of the points in the first point set are all less than or equal to a first preset threshold, the values of the points in the second point set are greater than the first preset threshold and less than the second preset threshold, the values of the points in the third point set are greater than or equal to the second preset threshold, and the second preset threshold is greater than the first preset threshold.
[0177] Optionally, the processing module 803 is further configured to:
[0178] Determine a difference between the first original feature candidate point set and the first point set to obtain a first intermediate set;
[0179] Determine the intersection of the first intermediate set and the second point set to obtain a second intermediate set;
[0180] A union of the second intermediate set and the third point set is determined to obtain a feature point candidate region of the second original image.
[0181] Optionally, the extraction module 804 is further configured to:
[0182] Determine multiple maximum points of the second original image according to the laser line scanning direction of the line laser scanning device, wherein each maximum point is a row-direction maximum point in each row or a column-direction maximum point in each column of the second original image, the row-direction maximum point is a feature point with the largest brightness in each row, and the column-direction maximum point is a feature point with the largest brightness in each column;
[0183] A feature candidate point set of the second original image is extracted based on the multiple maximum value points and each candidate feature point in the second original feature point candidate area.
[0184] Optionally, the extraction module 804 is further configured to:
[0185] Traversing each candidate feature point in the feature point candidate area of the second original image, comparing the currently traversed candidate feature point with the corresponding maximum value point, if the value of the currently traversed candidate feature point is greater than the value of the corresponding maximum value point, taking the currently traversed candidate feature point as a point in the feature candidate point set of the second original image;
[0186] Among them, if the maximum point corresponding to the currently traversed candidate feature point is a row-wise maximum point, then the maximum point corresponding to the currently traversed candidate feature point is the row-wise maximum point of the row where the currently traversed candidate feature point is located; if the maximum point corresponding to the currently traversed candidate feature point is a column-wise maximum point, then the maximum point corresponding to the currently traversed candidate feature point is the column-wise maximum point of the column where the currently traversed candidate feature point is located.
[0187] Optionally, the processing module 803 is further configured to:
[0188] Performing dilation processing on the feature point candidate region of the first differential image to obtain a processed feature point candidate region;
[0189] The extraction module 804 is further used for:
[0190] A feature candidate point set of the second original image is extracted from the second original image according to the processed feature point candidate region.
[0191] Optionally, the extraction module 804 is further configured to:
[0192] Determine multiple maximum points of the first original image according to the laser line scanning direction of the line laser scanning device, wherein each maximum point is a row-direction maximum point in each row or a column-direction maximum point in each column of the first original image, the row-direction maximum point is a feature point with the largest brightness value in each row, and the column-direction maximum point is a feature point with the largest brightness value in each column;
[0193] A plurality of maximum value points of the first original image are combined into a feature candidate point set of the first original image.
[0194] The above-mentioned device is used to execute the method provided by the aforementioned embodiment, and its implementation principle and technical effect are similar, which will not be repeated here.
[0195] The above modules may be one or more integrated circuits configured to implement the above methods, such as one or more application specific integrated circuits (ASIC), or one or more digital singnal processors (DSP), or one or more field programmable gate arrays (FPGA). For another example, when a module is implemented in the form of a processing element scheduling program code, the processing element may be a general-purpose processor, such as a central processing unit (CPU) or other processor that can call program code. For another example, these modules may be integrated together and implemented in the form of a system-on-a-chip (SOC).
[0196] Optionally, the present application also provides a program product, such as a computer-readable storage medium, including a program, which is used to execute the above method embodiment when executed by a processor.
[0197] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.
[0198] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0199] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional units.
[0200] The above-mentioned integrated unit implemented in the form of a software functional unit can be stored in a computer-readable storage medium. The above-mentioned software functional unit is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor (English: processor) to perform some steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (English: Read-Only Memory, abbreviated: ROM), random access memory (English: Random Access Memory, abbreviated: RAM), disk or optical disk and other media that can store program codes.
Claims
1. A method for extracting candidate feature points of an image, characterized in that: Applied to a processing device, the processing device is communicatively connected to a line laser scanning device, the method comprising: Acquire a compressed multi-frame laser scanning image sent by the line laser scanning device, wherein the multi-frame laser scanning image includes: a compressed first frame image, and a compressed multi-frame differential image, wherein the differential image is obtained from two adjacent frames of original images scanned by the line laser scanning device; Decompressing the compressed first frame image to obtain a first original image, taking the first original image as an initial first original image; and extracting a set of feature candidate points of the first original image; Decompressing the compressed multi-frame differential images to obtain at least one first differential image; determining a second original image according to the first original image and the first differential image; Dividing the first differential image into a plurality of point sets, each of the point sets includes points belonging to the same numerical interval; Determine a feature point candidate region of the second original image according to a feature candidate point set of the first original image and a plurality of point sets of the first differential image, wherein the second original feature point candidate region includes a plurality of candidate feature points; extracting a set of candidate feature points of the second original image from the second original image according to the candidate feature point region of the second original image; Determine whether the feature candidate point sets of the multiple frames of original images have been completely extracted. If not, use the second original image as the new first original image, use the adjacent differential image after the first differential image as the new first differential image, and re-execute the above steps of determining the second original image based on the first original image and the first differential image until the feature candidate point sets of the multiple frames of original images have been completely extracted.
2. The method according to claim 1, characterized in that The multiple point sets include: a first point set, a second point set, and a third point set. The values of the points in the first point set are all less than or equal to a first preset threshold, the values of the points in the second point set are greater than the first preset threshold and less than the second preset threshold, the values of the points in the third point set are greater than or equal to the second preset threshold, and the second preset threshold is greater than the first preset threshold.
3. The method according to claim 2, characterized in that The determining, based on the feature candidate point set of the first original image and the plurality of point sets of the first differential image, a feature point candidate region of the second original image comprises: Determine a difference between the first original feature candidate point set and the first point set to obtain a first intermediate set; Determine the intersection of the first intermediate set and the second point set to obtain a second intermediate set; A union of the second intermediate set and the third point set is determined to obtain a feature point candidate region of the second original image.
4. The method according to claim 1, characterized in that: The step of extracting a set of candidate feature points of the second original image from the second original image according to the candidate feature point region of the second original image comprises: Determine a plurality of maximum points of the second original image according to the laser line scanning direction of the line laser scanning device, wherein each maximum point is a row-direction maximum point in each row or a column-direction maximum point in each column of the second original image, the row-direction maximum point is a feature point with the largest brightness in each row, and the column-direction maximum point is a feature point with the largest brightness in each column; A feature candidate point set of the second original image is extracted based on the multiple maximum value points and each candidate feature point in the second original feature point candidate area.
5. The method according to claim 4, characterized in that The step of extracting a set of candidate feature points of the second original image according to the plurality of maximum points and each candidate feature point in the candidate feature point region of the second original image comprises: Traversing each candidate feature point in the feature point candidate area of the second original image, comparing the currently traversed candidate feature point with the corresponding maximum value point, and if the value of the currently traversed candidate feature point is greater than the value of the corresponding maximum value point, taking the currently traversed candidate feature point as a point in the feature candidate point set of the second original image; Among them, if the maximum point corresponding to the currently traversed candidate feature point is a row-wise maximum point, then the maximum point corresponding to the currently traversed candidate feature point is the row-wise maximum point of the row where the currently traversed candidate feature point is located; if the maximum point corresponding to the currently traversed candidate feature point is a column-wise maximum point, then the maximum point corresponding to the currently traversed candidate feature point is the column-wise maximum point of the column where the currently traversed candidate feature point is located.
6. The method according to claim 1, characterized in that Before extracting the set of candidate feature points of the second original image from the second original image according to the candidate feature point region of the second original image, the method further includes: Performing dilation processing on the feature point candidate region of the first differential image to obtain a processed feature point candidate region; The step of extracting a set of candidate feature points of the second original image from the second original image according to the candidate feature point region of the first differential image comprises: A feature candidate point set of the second original image is extracted from the second original image according to the processed feature point candidate area.
7. The method according to claim 1, characterized in that The step of extracting a set of candidate feature points from the first original image includes: Determine multiple maximum points of the first original image according to the laser line scanning direction of the line laser scanning device, wherein each maximum point is a row-direction maximum point in each row or a column-direction maximum point in each column of the first original image, the row-direction maximum point is a feature point with the maximum brightness value in each row, and the column-direction maximum point is a feature point with the maximum brightness value in each column; A plurality of maximum value points of the first original image are combined into a feature candidate point set of the first original image.
8. An image feature candidate point extraction device, characterized in that: Applied to a processing device, the processing device is communicatively connected with a line laser scanning device, and the device comprises: An acquisition module, used for acquiring a compressed multi-frame laser scanning image sent by the line laser scanning device, wherein the multi-frame laser scanning image includes: a compressed first frame image, and a compressed multi-frame differential image, wherein the differential image is obtained from two adjacent frames of original images scanned by the line laser scanning device; A decompression module, used for decompressing the compressed first frame image to obtain a first original image; A processing module, configured to use the first original image as an initial first original image; An extraction module, used to extract a set of feature candidate points of the first original image; The decompression module is further used to decompress the compressed multi-frame differential image to obtain at least one first differential image; The processing module is further used to determine a second original image according to the first original image and the first differential image; A division module, used for dividing the first differential image into a plurality of point sets, each of the point sets including points belonging to the same numerical interval; The processing module is further configured to determine a feature point candidate region of the second original image based on a feature point candidate set of the first original image and a plurality of point sets of the first differential image, wherein the second original feature point candidate region includes a plurality of candidate feature points; The extraction module is further used to extract a set of candidate feature points of the second original image from the second original image according to the candidate feature point region of the second original image; The processing module is also used to determine whether the feature candidate point sets of multiple frames of original images have been completely extracted. If not, the second original image is used as a new first original image, the adjacent differential image after the first differential image is used as a new first differential image, and the above-mentioned steps of determining the second original image based on the first original image and the first differential image are re-executed until the feature candidate point sets of multiple frames of original images are completely extracted.
9. A processing device, characterized in that: include: A processor, a storage medium and a bus, wherein the storage medium stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor communicates with the storage medium via the bus, and the processor executes the machine-readable instructions to perform the steps of any method as described in claims 1-7.
10. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are executed.
Citation Information
Patent Citations
Image transmission method and device for wireless three-dimensional scanner
CN111667436A
Blink detection device, blink detection program and blink detection method
JP2005004530A
System and method for splicing images
US20180018770A1
