A device and method for catenary image acquisition based on point cloud

Abstract

The application discloses a kind of based on point cloud's overhead line image acquisition device, it is related to overhead line detection technology, including vehicle body, the vehicle body is equipped with three-dimensional point cloud scanning module, walking mechanism and shooting unit, the three-dimensional point cloud scanning module is used to collect the three-dimensional point cloud data of the overhead line, the shooting unit includes suspension shooting unit, environment shooting unit and several hanging shooting unit, further include control system, the control system is used to obtain suspension position and hanging position according to three-dimensional point cloud data, when the vehicle body moves to suspension position, according to suspension position selects suspension shooting position, and according to suspension shooting position control the walking mechanism and the suspension shooting unit work, subsequently control suspension shooting unit collects suspension image, when vehicle body moves to hanging position, control hanging shooting unit collects hanging image, to the production demand of accurate detection of prior art to detect repair.

Description

Technical Field

[0001] This invention relates to the field of overhead contact line inspection technology, and in particular to an apparatus and method for acquiring overhead contact line images based on point clouds. Background Technology

[0002] The railway overhead contact system is the core power supply facility of electrified railways, consisting of contact wires, catenary wires, insulators, and support devices, responsible for transmitting 25kV high-voltage electricity to electric locomotives. Its operating status directly affects the stability of train power supply and train operation safety. Common faults include: mechanical defects, such as contact wire wear, broken strands, and loose suspension components; electrical faults, such as insulator flashover and overheating of electrical connectors; and environmental corrosion, such as rust on metal components and adhesion of dirt.

[0003] Traditional manual inspections rely on visual checks and simple tools (such as calipers and infrared thermometers), which suffer from low efficiency (≤3km per person per day), high risks associated with working at heights, and strong subjectivity in data recording. With the expansion of the high-speed rail network and the demand for intelligent upgrades, efficient and accurate automated inspection technology has become an inevitable trend.

[0004] In existing technologies, the main method for inspecting overhead contact lines is to use an overhead contact line suspension status detection and monitoring device (4C) to collect images of various components of the overhead contact line. However, this method often requires the device to be mounted on an inspection vehicle. Each inspection requires the cooperation of multiple departments and occupies the driving area. Inspections cannot be carried out during maintenance "windows". The inspection cycle is long and the timeliness is poor. At the same time, this method captures images of the overhead contact line suspension by detecting the presence of support pillars with a rangefinder. The camera takes pictures at a fixed angle. As a result, this method often requires more than twenty industrial cameras, which undoubtedly increases the difficulty of image analysis and reduces the accuracy of subsequent image recognition. Summary of the Invention

[0005] The purpose of this invention is to provide an apparatus and method for acquiring contact network images based on point clouds, so as to solve the above-mentioned problems.

[0006] This invention is achieved through the following technical solution:

[0007] A device for acquiring catenary images based on point clouds includes a vehicle body, on which a three-dimensional point cloud scanning module, a traveling mechanism, and a shooting unit are provided. The three-dimensional point cloud scanning module is used to acquire three-dimensional point cloud data of the catenary. The traveling mechanism is used to drive the vehicle body to move on the railway track. The shooting unit is used to acquire images of the catenary. The shooting unit is characterized by comprising a suspended shooting unit, an environmental shooting unit, and several suspended shooting units. Each suspended shooting unit is used to acquire suspended images of the catenary, and the image acquisition angles of each suspended shooting unit are different. The suspended shooting units are used to acquire suspended images of the catenary, and the environmental shooting units are used to acquire environmental images of the railway track.

[0008] It also includes a control system, which is used to control the working of the traveling mechanism according to the environmental image, and to obtain the suspension position and the suspended position according to the three-dimensional point cloud data. When the vehicle body moves to the suspension position, the suspension shooting position is selected according to the suspension position, and the traveling mechanism and the suspension shooting unit are controlled to work according to the suspension shooting position until the shooting focus of the suspension shooting unit coincides with the center of the suspension. Then, the suspension shooting unit is controlled to acquire the suspension image. When the vehicle body moves to the suspended position, the suspended shooting unit is controlled to acquire the suspended image.

[0009] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0010] This invention, through the design of the shooting unit, utilizes 3D point cloud data to obtain the position of the overhead contact line suspension and hanging, and changes the angle of the shooting unit to acquire images according to the position of the suspension and hanging to obtain corresponding images. Compared with the existing technology that uses fixed-angle shooting, this solution achieves multi-angle shooting by adjusting the angle of the shooting unit to acquire images, and has low requirements for the number of industrial cameras, which helps to reduce the difficulty of subsequent image analysis, improve the accuracy of image recognition, and reduce hardware costs.

[0011] Meanwhile, this solution utilizes 3D point cloud data to locate the suspension and hanging positions. Compared to the continuous shooting strategy used by existing monitoring devices, this solution requires fewer shots, which helps reduce data storage and analysis costs.

[0012] Furthermore, the environmental imaging unit includes a network camera, which is installed on the side wall of the vehicle body. The network camera is used to capture environmental images of the railway track, and the control system controls the operation of the traveling mechanism based on the environmental images.

[0013] Beneficial effects: Compared with existing technologies, this solution uses network cameras, which helps reduce the overall cost of the device. At the same time, network cameras have lower energy consumption, smaller size and higher cost performance, which helps reduce the device's energy consumption, weight and cost.

[0014] Furthermore, the suspended imaging unit includes a photosensitive sensor, a gimbal, and several first flashes. The photosensitive sensor, the gimbal, and the flashes are all mounted on the vehicle body. The photosensitive sensor is used to collect the ambient brightness, and the first flashes are used to adjust the ambient brightness. A first industrial camera is mounted on the gimbal, and the gimbal is used to drive the first industrial camera to rotate. The first industrial camera is used to collect suspended images. The control system controls the operation of the gimbal based on the three-dimensional point cloud data, and controls the exposure of the first industrial camera and the working duration of the first flashes based on the ambient brightness. At the same time, it controls the working power of the first flashes based on the ambient brightness.

[0015] Beneficial effects: The photosensitive sensor used in this solution collects the ambient brightness and collects the operating power of the first flash lamp based on the ambient brightness. Compared with the existing technology, this solution can reduce the reflection or artifacts in the image caused by excessive brightness of the first flash lamp, thereby improving the quality of the acquired image.

[0016] Furthermore, the suspended imaging unit includes a second industrial camera and a second flash, both of which are mounted on the vehicle body. The second industrial camera is used to acquire suspended images, and the second flash is used to adjust the ambient brightness. The control system controls the exposure of the second industrial camera and the working duration of the second flash based on the three-dimensional point cloud data.

[0017] Beneficial effects: By combining a second industrial camera and a second flash, this solution effectively avoids problems such as loss of detail caused by poor ambient light compared to existing technologies.

[0018] Furthermore, both the first industrial camera and the second industrial camera are equipped with telephoto lenses, which are used to change the focal length of the first industrial camera and the second industrial camera.

[0019] Beneficial effects: This solution, through the design of a telephoto lens, can magnify the image by using long focal length optics compared to existing technologies, effectively avoiding the problem of image detail loss caused by excessive distance between the vehicle body and the contact wire.

[0020] Furthermore, the traveling mechanism includes a frame on which a plurality of traveling wheels are mounted, and a servo planetary geared motor is installed in any one of the traveling wheels. The control system controls the operation of the servo planetary geared motor based on three-dimensional point cloud data and environmental images.

[0021] Beneficial effects: By designing the traveling wheels and servo planetary geared motors, this solution enables the vehicle body to move on the track compared to existing technologies, thereby adjusting the image acquisition angle and acquiring images from different locations on the overhead contact line. This helps reduce the number of cameras used, thus lowering the cost of the device.

[0022] Furthermore, the shooting unit also includes a supplementary light, which is equipped with an adjustment component. The adjustment component is used to change the angle between the optical axis of the supplementary light and the horizontal plane. The supplementary light is used to provide supplementary lighting for the suspension and hanging. The control system is also used to simultaneously acquire point cloud data at the same position and angle after acquiring the suspended image and the hanging image, select several feature points in the point cloud data, and acquire the positions of all feature points on the comparison suspended image or hanging image. The control system calculates the number of missing feature points in the suspended image or hanging image. When the number of missing feature points is greater than a set value, the control system will adjust the suspension... The image or suspended image is divided into several parts, and the ratio of the number of missing feature points in each part to the total number of feature points is calculated. The part with the largest ratio is selected as the supplementary lighting area. The control system controls the traveling mechanism to work again until the device moves to the position to acquire the suspended image or suspended image. Then, the adjustment component is controlled to work according to the supplementary lighting area until the optical axis of the supplementary light coincides with the supplementary lighting area. Then, the shooting unit and the supplementary light are controlled to work synchronously to acquire the suspended image or suspended image again, and the newly acquired suspended image or suspended image is overwritten with the previously acquired suspended image or suspended image.

[0023] Beneficial Effects: This solution assesses the completeness of information in the acquired suspended graphics and images by selecting schematic feature points in the 3D point cloud data. When the information in the image is insufficient, supplementary lighting is activated, and a second image is acquired at the corresponding location. Compared to existing technologies, this solution effectively avoids information loss caused by shadows from the first flash, second flash, or other ambient light sources, thus improving the accuracy of subsequent image analysis.

[0024] Furthermore, the control system is also used to adjust the operating power of the fill light according to the distance between the fill light area and the fill light.

[0025] Beneficial effects: By adjusting the working power of the supplementary light, this solution ensures that the brightness of the light is similar when it shines on different suspensions or hanging surfaces. Compared with existing technologies, this solution can help reduce poor lighting effects or contact wire reflections caused by insufficient light intensity, which affect image quality.

[0026] Furthermore, before selecting feature points, the control system performs grayscale processing on the suspended image or hanging image, selects the position with the smallest grayscale value after grayscale processing, constructs several concentric circles with the position with the smallest grayscale value as the center, and the number of feature points between adjacent concentric circles decreases as the radius of the concentric circles increases.

[0027] Beneficial effects: This solution performs grayscale processing on the image and roughly judges the shadow position based on the grayscale value, thereby selecting feature points. Compared with existing technologies, this solution can effectively improve the representativeness of feature points. At the same time, by focusing on verifying the positions with lower grayscale values, it can reduce the collection and processing of invalid data and improve the processing speed of this solution.

[0028] Furthermore, a method for acquiring contact network images based on point clouds, implemented using the aforementioned device, includes:

[0029] S1: Place the device on the railway where the data to be collected;

[0030] S2: Start the device, and at the same time, the user inputs the route to be collected into the control system;

[0031] S3: The device performs image acquisition and point cloud data acquisition, and transmits the point cloud data, environmental images, suspended images and hanging images back to the user;

[0032] S4: Based on the returned point cloud data, suspended images, and hanging images, the user determines whether the overhead contact line needs maintenance and the corresponding location, and adjusts the route to be collected based on the returned environmental images. Attached Figure Description

[0033] The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and form part of this application, do not constitute a limitation thereof. In the drawings:

[0034] Figure 1 This is a schematic diagram of the structure of the present invention;

[0035] Figure 2 This is a schematic diagram of the supplementary lighting portion in Embodiment 2 of the present invention;

[0036] Figure 3 This is a flowchart of the present invention.

[0037] The reference numerals in the attached figures represent: 1. Running wheel; 2. Vehicle body; 3. Suspended shooting unit; 31. Second industrial camera; 4. Suspended shooting unit; 41. Gimbal; 42. First industrial camera; 5. Environmental shooting unit; 6. Two-dimensional scanning radar; 7. Fill light; 8. Robotic arm; 9. Frame; 91. Fixed wheel frame; 92. Movable wheel frame. Detailed Implementation

[0038] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments and accompanying drawings. The illustrative embodiments and descriptions of this invention are for illustrative purposes only and are not intended to limit the invention. It should be noted that this invention is already in the actual research and development stage.

[0039] Example 1

[0040] like Figure 1 As shown, this embodiment includes: a vehicle body 2, on which a three-dimensional point cloud scanning module, a traveling mechanism, and a shooting unit are provided. The traveling mechanism is used to drive the vehicle body 2 to move on the railway track. The three-dimensional point cloud scanning module is used to collect three-dimensional point cloud data of the contact network. The three-dimensional point cloud scanning module includes a two-dimensional scanning radar 6 and an absolute encoder. The absolute encoder is used to collect the travel distance and travel speed of the device. The traveling mechanism includes a frame 9, which includes a fixed wheel frame 91 and a movable wheel frame 92. The fixed wheel frame 91 is fixedly connected to the side wall of the vehicle body 2 by bolts. The movable wheel frame 92 is fixedly connected to the side wall away from the fixed wheel frame 91 by bolts. Several traveling wheels 1 are installed on the side of the fixed wheel frame 91 and the movable wheel frame 92 away from the vehicle body 2. A servo planetary geared motor is installed in any one of the traveling wheels 1, and the absolute encoder is installed on any one of the traveling wheels 1 on the same side as the servo planetary geared motor.

[0041] The shooting unit is used to collect images of the overhead contact line. The shooting unit includes a suspended shooting unit 4, an environmental shooting unit 5, and several suspended shooting units 3. The suspended shooting units 3 are respectively arranged on both sides of the top wall of the vehicle body 2.

[0042] The suspended imaging units 3 are all used to collect suspended images of the overhead contact line, and the image acquisition angles of the suspended imaging units are all different. The suspended imaging unit 3 includes a second industrial camera 31 and a second flash. The second industrial camera 31 and the second flash are both installed on the top wall of the vehicle body 2. The second industrial camera 31 is used to collect suspended images, and the second flash is used to adjust the ambient brightness.

[0043] The suspended imaging unit 4 is used to acquire images of the overhead contact line. The suspended imaging unit 4 includes a photosensitive sensor, a gimbal 41, and several first flash lamps. The photosensitive sensor, the gimbal 41, and the flash lamps are all mounted on the vehicle body 2. The photosensitive sensor is used to acquire the ambient brightness, and the flash lamps are used to adjust the ambient brightness. A first industrial camera 42 is mounted on the gimbal 41, and a first telephoto lens is mounted on the first industrial camera 42. The gimbal 41 is used to drive the first industrial camera 42 to rotate, and the first industrial camera 42 is used to acquire suspended images.

[0044] In this embodiment, both the first industrial camera 42 and the second industrial camera 31 have a resolution of 6500W pixels, and both use 50mm telephoto lenses.

[0045] The environmental imaging unit 5 is used to collect environmental images of the railway track. The environmental imaging unit 5 includes a network camera, which is installed on the side wall of the car body 2. The network camera is used to collect environmental images of the railway track.

[0046] It also includes a control system. The control system acquires three-dimensional point cloud data based on the two-dimensional scanning radar 6, the device's travel distance, and travel speed. The control system controls the operation of the traveling mechanism based on the environmental image and acquires the suspension position and hanger position based on the three-dimensional point cloud data. When the vehicle body 2 moves to the hanger position, it selects the hanger shooting position based on the hanger position and controls the traveling mechanism and the hanger shooting unit 4 to operate until the shooting focus of the hanger shooting unit 4 coincides with the center of the hanger. Then, it controls the hanger shooting unit 4 to acquire hanger images. During the operation of the hanger shooting unit 4, the control system controls the operation based on the three-dimensional point cloud data. The control system operates the gimbal 41 and controls the exposure of the first industrial camera 42 and the working duration of the first flash lamp according to the ambient brightness. At the same time, it controls the working power of the first flash lamp according to the ambient brightness. When the vehicle body 2 moves to the suspended position, it controls the suspended shooting unit 3 to acquire the suspended image. During this process, the control system controls the exposure of the second industrial camera 31 and the working duration of the second flash lamp according to the three-dimensional point cloud data. The control system includes an embedded host, and the first industrial camera 42, the second industrial camera 31, the first flash lamp, the second flash lamp, the absolute encoder, the servo planetary gear motor, the photosensitive sensor, the gimbal 41, and the network camera are all electrically connected to the embedded host.

[0047] Detailed implementation: During the use of this device, the device is placed on the railway track and started. The network camera continuously collects environmental images around the track. The embedded host performs feature recognition based on the environmental images to determine the direction of travel of the device. Then, based on the determination of the direction of travel of the device, the servo planetary gear motor is controlled to work. The servo planetary gear motor drives the traveling wheel 1 to move, thereby driving the device to move on the track.

[0048] Simultaneously, the absolute encoder synchronously collects the distance and speed traveled by the device during its movement, while the 2D scanning radar 6 continuously scans the 2D profile point cloud data along the direction of travel. This 2D profile point cloud data is then stitched together with the device's travel distance and speed data to obtain 3D point cloud data. During this process, based on the acquired 3D point cloud data, the system analyzes in real time whether the scanned components of the overhead contact line are suspended. If the components are not suspended, the embedded host continues to control the servo planetary geared motor, allowing the device to continue moving along the railway.

[0049] When a component is identified as suspended, the embedded host controls the second industrial camera 31 and the second flash to work after a certain delay. The delay time is determined by the embedded host based on the distance between the two-dimensional scanning radar 6 and the second industrial camera 31, the device's travel speed, and the time required to identify whether the component is suspended. The delay time is used to ensure that the focus of the second industrial camera 31 is at the center of the suspension when it is working, so that the suspension is located in the exact center of the acquired suspension image, which helps to increase the amount of information that can be covered in the acquired image.

[0050] Meanwhile, during the device's movement, the embedded host continuously determines whether the scanned component is suspended based on the 3D point cloud data. When the component is suspended, the embedded host controls the servo planetary geared motor to move backward in the opposite direction of travel until the device moves to the rear of the suspension. It then controls the pan-tilt unit 41 to operate and position the first industrial camera 42 at the optimal shooting angle. The optimal position is determined by the focal distance of the first industrial camera 42, the spatial position of the suspension, the pitch angles of the first industrial camera 42, and the rotation angle of the first industrial camera 42. When selecting the optimal position, the embedded host needs to ensure that the focal point of the first industrial camera 42 coincides with the center position of the suspension, so as to ensure that the suspension is centered in the acquired suspension image.

[0051] The embedded host then performs feature extraction and real-time analysis on the acquired hanging image. Based on the acquired hanging features, it determines whether the hanging image completely covers the hanging structure. If the hanging image cannot completely cover the hanging structure, the host controls the pan-tilt unit 41 to work again, adjusting the shooting angle of the first industrial camera 42, and then controls the first industrial camera 42 to work again. The embedded host repeats the above steps, adjusting the shooting angle of the first industrial camera 42, until all information about the hanging structure at that position is acquired.

[0052] The embedded host then uses a servo planetary geared motor to drive the device to the front of the suspension, repeating the above steps until all information about the suspension at that position is acquired. This achieves the acquisition of suspension information.

[0053] The embedded host sends suspension and hanging information to the user, who can then judge the status of suspension and hanging based on this information.

[0054] Example 2

[0055] As attached Figure 2 As shown, the difference from the above embodiment is that the shooting unit further includes a fill light 7, and the fill light 7 is also equipped with an adjustment component. In this embodiment, the adjustment component includes a robotic arm 8, which is mounted on the vehicle body 2, and the fill light 7 and the robotic arm 8 are detachably connected by a buckle. The adjustment component is used to change the angle between the optical axis of the fill light 7 and the horizontal plane, and the fill light 7 is used to provide supplementary lighting for the suspension and the hanging.

[0056] Both the robotic arm 8 and the fill light 7 are electrically connected to the embedded host.

[0057] The embedded host is also used to simultaneously acquire point cloud data at the same position and angle (i.e., the position and angle of the suspended or suspended image) after acquiring the suspended image and the hanging image. Several feature points are selected from the point cloud data. When selecting feature points, the embedded host first performs grayscale processing on the suspended or suspended image, selects the position with the smallest grayscale value after grayscale processing, and constructs several concentric circles with the position with the smallest grayscale value as the center. The number of feature points between adjacent concentric circles decreases as the radius of the concentric circles increases. Then, the positions of all feature points on the suspended or suspended image are acquired, and the number of missing feature points in the suspended or suspended image is calculated.

[0058] When the number of missing feature points is less than or equal to the set value, the embedded host controls the servo planetary gear motor to continue working and carry out subsequent image acquisition of the overhead contact line on the track.

[0059] When the number of missing feature points exceeds a set value, the control system divides the suspended image or hanging image into several parts and calculates the ratio of the number of missing feature points in each part to the number of selected feature points. The part with the largest ratio of missing feature points to selected feature points is selected as the supplementary lighting area. The embedded host controls the servo planetary geared motor to work again until the device moves to the position where the suspended image or hanging image was last acquired. Then, according to the supplementary lighting area, the robotic arm 8 is controlled to work and adjust the angle of the optical axis of the supplementary light 7 until the optical axis of the supplementary light 7 coincides with the supplementary lighting area. Then, the first industrial camera 42, the second industrial camera 31, the first flash, the second flash, and the supplementary light 7 are controlled to work simultaneously to acquire the suspended image or hanging image again, and the newly acquired suspended image or hanging image is overwritten with the previously acquired suspended image or hanging image.

[0060] The embedded host is also used to control the operating power of the fill light 7 based on the distance between the fill light area and the fill light 7.

[0061] The specific operating steps are as follows: During the image acquisition process using the device, due to factors such as the angle of the first flash, the second flash, or ambient light, shadows may inevitably appear on the acquired suspended and hanging images. When the above situation occurs, it is easy to cause data loss in the acquired images, affecting the operator's judgment of the contact wire status.

[0062] Therefore, after acquiring a suspended or hanging image at a certain location, the embedded host verifies the integrity of the acquired suspended or hanging image.

[0063] During the judgment, the embedded host obtains point cloud data at the position and angle of the first industrial camera 42 or the second industrial camera 31 when the image of the suspended or hanging object to be verified is captured, based on the already acquired 3D point cloud data.

[0064] Simultaneously, the embedded host performs grayscale processing on the suspended or hanging image to be verified. Since the brightness of the shadow area of ​​the image is low, after grayscale conversion, the grayscale value at that location is usually lower than that of the surrounding non-shadow areas. Therefore, the location with the lowest grayscale value is very likely to be a shadow area. In this case, the location is selected as the key verification location. Using this location as a circle, several concentric circles are constructed. When selecting feature points, the number of feature points between adjacent concentric circles gradually decreases as the radius of the concentric circles increases, so that the feature points are distributed in a state of diffusion from the circle to the surrounding areas. Compared with the scheme of using uniform feature point selection, the feature point selection in this scheme has a certain degree of randomness, which improves the representativeness and reliability of the selected feature points. At the same time, this scheme uses the location that may be a shadow area for key verification, thereby effectively reducing the amount of computation in invalid areas and helping to reduce computational costs.

[0065] Feature points are points that are significantly different from other locations, such as corners, inflection points, and points on the edges of objects.

[0066] Once the feature points are selected, the embedded host predicts their corresponding positions in the suspended or hanging image based on their positions in the point cloud data. Then, it identifies the area around the predicted position. If there is no shadow occlusion or the shadow occlusion does not cause information loss at that position, then there will be a relatively obvious difference in brightness or color around the predicted position during the identification process. At this time, it can be determined that the feature point exists at that position. If there is no obvious difference in brightness or color around the predicted position, that is, the feature point cannot be located by brightness or color difference, then it is determined that the feature point is missing.

[0067] Subsequently, the suspended or hanging image is divided into several parts. For example, if the size of image A is 10*10, it is divided into 100 images of size 1*1. The ratio of the number of missing feature points to the number of selected feature points in each image is calculated. If the ratio is small, the corresponding image has less missing information, and it can be inferred that there is no shadow or the shadow has little influence at the corresponding position of the image.

[0068] Conversely, if the ratio is large, the corresponding image has more missing information. In this case, the embedded host uses the position corresponding to that image as the fill light area. For example, in the image A, which is divided into equal parts, the image in the first column of the sixth row has the largest ratio. Therefore, the position corresponding to the first column of the sixth row is selected as the fill light area. The embedded host uses a servo planetary geared motor to drive the device back to the position where the image was acquired. At the same time, the embedded host controls the robotic arm 8 to work according to the fill light area. Through the work of the robotic arm 8, the optical axis of the fill light 7 is aligned with the center of the fill light area. Then, the embedded host repeats the image acquisition steps, that is, it controls the suspended shooting unit 3 or the hanging shooting unit 4 to work again and simultaneously activates the fill light 7 to acquire the image at that position again, and overwrites the previously acquired image with missing information with the image acquired this time. Furthermore, the working power of the fill light 7 is determined by the straight-line distance between the vehicle body 2 and the fill light area, the angle between the fill light 7 and the fill light area, and the brightness of the environment. By dynamically adjusting the working power of the fill light 7, the negative impact on image quality caused by insufficient brightness of the fill light 7, poor fill light effect, and excessive brightness of the fill light 7 causing reflection or artifacts in the fill light area can be effectively reduced.

[0069] Compared to existing technologies, this solution verifies newly acquired images to determine the impact of shadows cast by ambient light sources or flash on image quality. When shadows significantly affect image quality, supplementary lighting and secondary image acquisition are performed, thereby increasing the amount of information contained in the acquired images throughout the process and reducing the possibility of operators missing or misjudging contact wire damage due to shadow obstruction.

[0070] Example 3

[0071] As attached Figure 3 As shown, the difference from the above embodiments is that the method for acquiring contact network images based on point clouds, implemented using the above-described device, includes:

[0072] S1: Place the device on the railway where the data to be collected;

[0073] S2: Start the device, and at the same time, the user inputs the route to be collected into the control system;

[0074] S3: The device travels along the route to be collected, and collects images and point cloud data along the way, and transmits the point cloud data, environmental images, suspended images and hanging images back to the user.

[0075] S4: The user determines whether the overhead contact line needs maintenance based on the returned point cloud data, suspended images, and hanging images, and at the location where maintenance is needed, while also adjusting the route to be collected based on the returned environmental images.

[0076] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A device for acquiring contact network images based on point clouds, comprising a vehicle body (2), wherein the vehicle body (2) is provided with a three-dimensional point cloud scanning module, a traveling mechanism, and a shooting unit, wherein the three-dimensional point cloud scanning module is used to acquire three-dimensional point cloud data of the contact network, the traveling mechanism is used to drive the vehicle body (2) to move on the railway track, and the shooting unit is used to acquire images of the contact network, characterized in that: The shooting unit includes a suspended shooting unit (4), an environmental shooting unit (5), and several suspended shooting units (3). The suspended shooting units (3) are all used to collect suspended images of the contact wire, and the image collection angles of the suspended shooting units (3) are all different. The suspended shooting unit (4) is used to collect suspended images of the contact wire, and the environmental shooting unit (5) is used to collect environmental images of the railway track. It also includes a control system, which is used to control the working of the traveling mechanism according to the environmental image, and to obtain the suspension position and the suspended position according to the three-dimensional point cloud data. When the vehicle body (2) moves to the suspension position, the suspension shooting position is selected according to the suspension position, and the traveling mechanism and the suspension shooting unit (4) are controlled to work according to the suspension shooting position until the shooting focus of the suspension shooting unit (4) coincides with the center of the suspension. Then the suspension shooting unit (4) is controlled to collect the suspension image. When the vehicle body (2) moves to the suspended position, the suspension shooting unit (3) is controlled to collect the suspension image. The shooting unit also includes a fill light (7), which is equipped with an adjustment component. The adjustment component is used to change the angle between the optical axis of the fill light (7) and the horizontal plane. The fill light (7) is used to provide fill light for the suspension and hanging. The control system is also used to simultaneously acquire point cloud data at the same position and angle after acquiring the suspended image and the hanging image, select several feature points in the point cloud data, perform grayscale processing on the suspended image or hanging image when selecting feature points, select the position with the smallest grayscale value after grayscale processing in the image, construct several concentric circles with the position with the smallest grayscale value as the center, and the number of feature points between adjacent concentric circles decreases as the radius of the concentric circles increases. Then, the positions of all feature points on the suspended image or hanging image are acquired. The number of missing feature points in the suspended image or hanging image is calculated. When the number of missing feature points is greater than a set value, the control system divides the suspended image or hanging image into several parts and calculates the ratio of the number of missing feature points in each part to the total number of feature points. The part with the largest ratio is selected as the supplementary lighting area. The control system controls the traveling mechanism to work again until the device moves to the position to acquire the suspended image or hanging image. Then, the adjustment component is controlled to work according to the supplementary lighting area until the optical axis of the supplementary light (7) coincides with the supplementary lighting area. Then, the shooting unit and the supplementary light (7) are controlled to work synchronously to acquire the suspended image or hanging image again and overwrite the previously acquired suspended image or hanging image with the newly acquired suspended image or hanging image. The control system is also used to adjust the working power of the fill light (7) according to the distance between the fill light area and the fill light (7).

2. The device for acquiring contact network images based on point clouds according to claim 1, characterized in that: The environmental imaging unit (5) includes a network camera, which is installed on the side wall of the vehicle body (2). The network camera is used to collect environmental images of the railway track, and the control system controls the working of the traveling mechanism according to the environmental images.

3. The device for acquiring contact network images based on point clouds according to claim 1, characterized in that: The suspended shooting unit (4) includes a photosensitive sensor, a gimbal (41), and several first flash lamps. The photosensitive sensor, the gimbal (41), and the flash lamps are all mounted on the vehicle body (2). The photosensitive sensor is used to collect the brightness of the environment, and the first flash lamps are used to adjust the brightness of the environment. The gimbal (41) is equipped with a first industrial camera (42). The gimbal (41) is used to drive the first industrial camera (42) to rotate. The first industrial camera (42) is used to collect suspended images. The control system controls the operation of the gimbal (41) according to the three-dimensional point cloud data, and controls the exposure of the first industrial camera (42) and the working duration of the first flash lamps according to the brightness of the environment. At the same time, it controls the working power of the first flash lamps according to the brightness of the environment.

4. The device for acquiring contact network images based on point clouds according to claim 3, characterized in that: The suspended shooting unit (3) includes a second industrial camera (31) and a second flash. The second industrial camera (31) and the second flash are both mounted on the vehicle body (2). The second industrial camera (31) is used to acquire suspended images, and the second flash is used to adjust the ambient brightness. The control system controls the exposure of the second industrial camera (31) and the working time of the second flash according to the three-dimensional point cloud data.

5. The device for acquiring contact network images based on point clouds according to claim 4, characterized in that: Both the first industrial camera (42) and the second industrial camera (31) are equipped with telephoto lenses, which are used to change the focal length of the first industrial camera (42) and the second industrial camera (31).

6. The device for acquiring contact network images based on point clouds according to claim 1, characterized in that: The traveling mechanism includes a frame (9), on which a plurality of traveling wheels (1) are mounted. A servo planetary geared motor is installed in any one of the traveling wheels (1). The control system controls the servo planetary geared motor to work based on three-dimensional point cloud data and environmental images.

7. A method for acquiring contact wire images based on point clouds, implemented using the device for acquiring contact wire images based on point clouds according to any one of claims 1-6, characterized in that: include: S1: Place the device on the railway where the data to be collected; S2: Start the device, and at the same time, the user inputs the route to be collected into the control system; S3: The device performs image acquisition and 3D point cloud data acquisition, and transmits the point cloud data, environmental images, suspended images and hanging images back to the user; S4: Perform feature recognition on the transmitted suspended and hanging images to determine whether the contact wire needs maintenance and its corresponding location, and adjust the route to be collected based on the transmitted environmental images.

Images