CN-121998813-A - Tunnel inner wall image unfolding method, device, terminal, medium and program product

CN121998813ACN 121998813 ACN121998813 ACN 121998813ACN-121998813-A

Abstract

The invention relates to the technical field of image processing and provides a method, a device, a terminal, a storage medium and a computer program product for expanding images of the inner wall of a tunnel, wherein the method is applied to two-dimensional plane expansion of sequence images of the inner wall of the tunnel acquired by an array camera system, the array camera system is provided with n+1 cameras which are sequentially arranged in a circumferential array according to sequence numbers, and n is a positive integer; the method for unfolding the tunnel inner wall image comprises the steps of obtaining the tunnel inner wall sequence image acquired by the array camera system, and carrying out rigid body transformation on the tunnel inner wall sequence image by utilizing a preset rigid body transformation model to obtain a two-dimensional plane panorama of the tunnel inner wall sequence image. According to the scheme, the tunnel inner wall sequence image acquired by the array camera system can be unfolded into the two-dimensional plane panorama which is free of distortion, measurable and lossless in resolution through the full rigid body transformation process, any three-dimensional reconstruction or laser point cloud is not needed, and the cost is low and the resolution is high.

Inventors

CHEN PAN
CAO DONG
ZHOU LANGMING
YIN LUBING
CHEN LEI
YUE LEI

Assignees

湖南科天健光电技术有限公司

Dates

Publication Date: 20260508
Application Date: 20260122

Claims (10)

1. The method is characterized by being applied to two-dimensional plane expansion of tunnel inner wall sequence images acquired by an array camera system, wherein the array camera system is provided with n+1 cameras which are sequentially arranged in a circumferential array according to serial numbers, and n is a positive integer, and the method comprises the following steps: Acquiring the sequence images of the inner wall of the tunnel acquired by the array camera system; And carrying out rigid transformation on the tunnel inner wall sequence image by using a preset rigid transformation model to obtain a two-dimensional plane panorama of the tunnel inner wall sequence image.
2. The method for developing an image of an inner wall of a tunnel according to claim 1, wherein the rigid transformation model set in advance comprises a hoop rigid transformation model and a longitudinal affine transformation model; Performing rigid transformation on the tunnel inner wall sequence image by using a preset rigid transformation model to obtain a two-dimensional plane panorama of the tunnel inner wall sequence image, wherein the method comprises the following steps: performing annular rigid transformation on the tunnel inner wall sequence image by using the annular rigid transformation model to obtain a transverse expansion diagram of the tunnel inner wall sequence image; Performing longitudinal affine transformation on the transverse expansion map of the tunnel inner wall sequence image by using the longitudinal affine transformation model to obtain a global expansion map of the tunnel inner wall sequence image; And taking the global expansion diagram of the tunnel inner wall sequence image as a two-dimensional plane panorama of the tunnel inner wall sequence image.
3. The method for developing an image of an inner wall of a tunnel according to claim 2, wherein performing the hoop rigid transformation on the sequence of images of the inner wall of the tunnel using the hoop rigid transformation model to obtain a transverse developed view of the sequence of images of the inner wall of the tunnel, comprises: Extracting features of preset transverse key points by using the annular rigid body transformation model aiming at an overlapping region of the tunnel inner wall images acquired by cameras with adjacent serial numbers in the tunnel inner wall sequence images; estimating the rotation angle and the translation amount required by rigid transformation of the camera images with adjacent serial numbers according to the extracted characteristics of the preset transverse key points; Performing rigid transformation on the camera images with adjacent serial numbers according to the rotation angle and the translation amount obtained by estimation to obtain a 2D rigid transformation matrix as a local rigid transformation matrix; Performing chain accumulation on a reference camera coordinate system in the array camera system aiming at local rigid body transformation matrixes of all adjacent camera images in the array camera system to obtain global rigid body transformation matrixes of all camera images in the array camera system; And mapping pixels of the corresponding images to a transverse unfolding plane aiming at the global rigid body transformation matrix of all camera images in the array camera system, and carrying out linear feathering fusion on overlapping differences to obtain a transverse unfolding diagram of the tunnel inner wall sequence image.
4. A tunnel inner wall image development method according to claim 2 or 3, wherein performing longitudinal affine transformation on the transverse development view of the tunnel inner wall sequence image by using the longitudinal affine transformation model to obtain a global development view of the tunnel inner wall sequence image comprises: Extracting features of preset longitudinal key points by using the longitudinal affine transformation model according to the overlapping region of adjacent frames of the transverse expansion graph of the tunnel inner wall sequence image; estimating the rotation angle and the translation amount required by rigid transformation of the images of the adjacent frames according to the extracted characteristics of the preset longitudinal key points; performing rigid transformation on the images of the adjacent frames according to the rotation angle and the translation amount obtained by estimation to obtain a 2D rigid transformation matrix as a local rigid transformation matrix, thereby obtaining local rigid transformation matrices of the images of all the adjacent frames in the transverse expansion diagram of the tunnel inner wall sequence images; Aiming at the local rigid body transformation matrix of the images of all adjacent frames in the transverse expansion diagram of the tunnel inner wall sequence image, carrying out chain accumulation to a first frame coordinate system in the transverse expansion diagram of the tunnel inner wall sequence image to obtain an image global rigid body transformation matrix of all frames in the transverse expansion diagram of the tunnel inner wall sequence image; And mapping pixels of the corresponding images to a transverse unfolding plane aiming at the image global rigid body transformation matrix of all frames in the transverse unfolding diagram of the tunnel inner wall sequence image, and carrying out linear eclosion fusion on overlapping differences to obtain the global unfolding diagram of the tunnel inner wall sequence image.
5. The tunnel inner wall image development method according to any one of claims 1 to 4, characterized by further comprising: And performing flow slicing on the global expansion map of the tunnel inner wall sequence image by using a preset block slicing flow frame to obtain an image sequence of the two-dimensional plane panorama of the tunnel inner wall sequence image.
6. The method for expanding an image of a tunnel inner wall according to claim 5, wherein the step of performing a stream slicing on the global expanded view of the sequence of tunnel inner wall images to obtain an image sequence of a two-dimensional planar panorama of the sequence of tunnel inner wall images comprises: In the process of outputting the global expansion map of the tunnel inner wall sequence image, determining whether the height of the global expansion map of the tunnel inner wall sequence image is larger than or equal to a preset cutting value; if the height of the global expansion map of the tunnel inner wall sequence image is determined to be larger than or equal to a preset cutting value, executing a preset slicing action to obtain a current slicing map; Performing slice position calculation on a coordinate chain for performing rigid transformation on the tunnel inner wall sequence image, and recording the current slice according to the calculated slice position; And obtaining all slice images of the two-dimensional plane panorama of the tunnel inner wall sequence image, and forming an image sequence of the two-dimensional plane panorama of the tunnel inner wall sequence image.
7. A tunnel inner wall image development device for realizing the development of a tunnel inner wall image by using the tunnel inner wall image development method according to any one of claims 1 to 6, comprising: an acquisition unit configured to acquire the tunnel inner wall sequence image acquired by the array camera system; and the control unit is configured to perform rigid transformation on the tunnel inner wall sequence image by using a preset rigid transformation model to obtain a two-dimensional plane panorama of the tunnel inner wall sequence image.
8. A terminal, comprising the tunnel inner wall image developing device according to claim 7.
9. A storage medium comprising a stored program, wherein the program, when run, controls a device in which the storage medium is located to perform the tunnel inner wall image development method of any one of claims 1 to 6.
10. A computer program product comprising a computer program, characterized in that the computer program, when being executed by a processor, implements the steps of the tunnel inner wall image development method of any one of claims 1 to 6.

Description

Tunnel inner wall image unfolding method, device, terminal, medium and program product Technical Field The invention relates to the technical field of image processing, in particular to a tunnel inner wall image unfolding method, a device, a terminal, a storage medium and a computer program product, and especially relates to a tunnel inner wall high-definition image plane unfolding method, a device, a terminal, a storage medium and a computer program product based on rigid body transformation. Background The panoramic image of the inner wall of the tunnel is widely used for tasks such as health monitoring, crack detection, water leakage identification and the like of the tunnel structure. Because the tunnel is a cylinder or a cylinder-like curved surface, the acquired high-definition image sequence has geometric distortion, and the high-definition image sequence can be measured and identified only after being unfolded into a two-dimensional plane. However, in the related scheme, the high-definition image sequence of the inner wall of the tunnel is unfolded to be a two-dimensional plane, three-dimensional reconstruction or laser point cloud is needed, the cost is high, and the resolution is lost. Therefore, development of a method, an apparatus, a terminal, a storage medium and a computer program product for developing an image of a tunnel inner wall is needed, and in particular, relates to a method, an apparatus, a terminal, a storage medium and a computer program product for developing an image of a tunnel inner wall, and more particularly, to a method, an apparatus, a terminal, a storage medium and a computer program product for developing a high-definition image plane of a tunnel inner wall based on rigid transformation, which can develop a sequence image of a tunnel inner wall acquired by an array camera system into a distortion-free, measurable and resolution-free two-dimensional planar panorama at one time through an all-rigid transformation process, without any three-dimensional reconstruction or laser point cloud, with low cost and high resolution. The foregoing is provided merely for the purpose of facilitating understanding of the technical solutions of the present invention and is not intended to represent an admission that the foregoing is prior art. Disclosure of Invention The invention aims to provide a tunnel inner wall image unfolding method, a device, a terminal, a storage medium and a computer program product, in particular to a tunnel inner wall high-definition image plane unfolding method, a device, a terminal, a storage medium and a computer program product based on rigid body transformation, which can unfold a tunnel inner wall sequence image acquired by an array camera system into a two-dimensional plane panorama without distortion, measurability and lossless resolution through a full rigid body transformation process without any three-dimensional reconstruction or laser point cloud, and have low cost and high resolution. In order to solve the technical problems, as one aspect of the invention, a method for expanding a tunnel inner wall image is provided, which is applied to two-dimensional plane expansion of a tunnel inner wall sequence image acquired by an array camera system, wherein the array camera system is provided with n+1 cameras which are sequentially arranged in a circumferential array according to serial numbers, n is a positive integer, the method for expanding the tunnel inner wall image comprises the steps of acquiring the tunnel inner wall sequence image acquired by the array camera system, and performing rigid body transformation on the tunnel inner wall sequence image by utilizing a preset rigid body transformation model to obtain a two-dimensional plane panorama of the tunnel inner wall sequence image. According to an example embodiment of the invention, the preset rigid transformation model comprises a circumferential rigid transformation model and a longitudinal affine transformation model, wherein the preset rigid transformation model is utilized to conduct rigid transformation on the tunnel inner wall sequence image to obtain a two-dimensional plane panorama of the tunnel inner wall sequence image, the preset rigid transformation model comprises the step of conducting circumferential rigid transformation on the tunnel inner wall sequence image by utilizing the circumferential rigid transformation model to obtain a transverse expansion map of the tunnel inner wall sequence image, the step of conducting longitudinal affine transformation on the transverse expansion map of the tunnel inner wall sequence image by utilizing the longitudinal affine transformation model to obtain a global expansion map of the tunnel inner wall sequence image, and the step of taking the global expansion map of the tunnel inner wall sequence image as the two-dimensional plane panorama of the tunnel inner wall sequence image. According to an example embodiment of the invention, the circum