1. bookAHEAD OF PRINT
Détails du magazine
License
Format
Magazine
eISSN
2444-8656
Première parution
01 Jan 2016
Périodicité
2 fois par an
Langues
Anglais
Accès libre

Research on non-linear visual matching model under inherent constraints of images

Publié en ligne: 20 May 2022
Volume & Edition: AHEAD OF PRINT
Pages: -
Reçu: 22 Mar 2022
Accepté: 10 Apr 2022
Détails du magazine
License
Format
Magazine
eISSN
2444-8656
Première parution
01 Jan 2016
Périodicité
2 fois par an
Langues
Anglais
Introduction

As one of the research hotspots of computer vision, visual matching has the preponderance of estimating the 3D structure and appearance information of the scene. The advantages of convenience and low cost of this technology promote the development of artificial intelligence, pattern recognition and machine vision. However, with the development of industrial refinement, people put forward higher requirements for the accuracy and range of measurement information [1, 2]. For example, for accurate identification of workpiece position, precise data are needed to calculate centroid coordinates on this basis, which introduces the need for higher measurement accuracy. At the same time, most of these devices are foreign products, which are expensive, suitable for customisation and detection of specific scenes, and not universal, which severely restricts the development of China's manufacturing industry [3,4,5]. In addition, with the continuous development of visual matching techniques, the single-scale application scenarios represented by mobile phones will greatly increase. At the same time, among the many ways of visual model reconstruction, 3D reconstruction technology, which is based on stereo vision matching, utilises the bionics principle. The acquired images are processed by imitating human visual perception, which makes the effect relatively close to the recognition result of human eyes, and endows it with the capability to flexibly process data information in complex scenes. Therefore, this function cannot be replaced by other computer vision methods, and it is also the general trend of future development.

Extraction and matching of a feature point is an important step in determining whether the result of visual matching is accurate or not, and is also an important cornerstone to ensure the smooth development of the follow-up work. Among the existing 3D reconstruction methods, SIFT fast feature matching algorithm, SURF accelerated robust feature matching algorithm and ORB algorithm are commonly used [6, 7]. But the real-time performance of SIFT is poor, even after the underlying optimisation, through several times improvement, the time performance is not enough to meet the real-time requirements of current application scenarios. In the application scenario of scale-free variation, if SURF algorithm is used, the time cost will increase, while if ORB algorithm is used, the accuracy will decrease [8, 9]. Therefore, for the inherent constraint problem of images in the scale-free variation scene, we need to find a better way to solve the balance between visual matching effect and speed from the extraction, matching and misjudgement point elimination of feature point.

Establishment of non-linear visual matching model
Principle of visual imaging

The environment of visual imaging process consists of scene illumination, light propagation, illumination of image plane and signal processing. Natural scenery is ever-changing with different shapes. It is necessary to know the background characteristics of various radiation sources and targets, whether the scenery can be imaged and with what means it is suitable to be converted into an image. First, most of the images used in image processing are collected according to the principle of reflected light imaging. In the design of actual imaging system, it is generally simplified to Lambert model, and the ideal diffuse reflector is the typical Lambert [10], where the scenery emits light in all directions. Take a point in the scene as an example to analyse the light propagation. Figure 1 is a sectional view of an imaging system, and point P emits spherical waves; when it reaches the lens system, due to the differences between points on the surface of spherical waves and the refractive index of the lens, the light rays converge to point P′ of the image plane. Due to the coaxial sphericity of the imaging system, the object point P corresponds to the only image point P′. When the spherical waves emitted by all points in the field of view can be received by the imaging system, a two-dimensional image of the scene is formed.

Fig. 1

Cross-sectional view of visual imaging system

As shown in Figure 2, the light received by the human eye reaches the fundus, and passes through the human eye imaging and the eyepiece after reflection, thereby forming an image at plane S behind the eyepiece. Then, the light passes through the reducing glass, and the common objective glass receives light rays where the plane S becomes the common focal plane of the eyepiece and the two lenses behind it. Afterwards, the imaging beam forms two parallel beams after sharing the objective lens, and the distance between the two parallel beams is increased by the reflector group, which provides space for adding mirror of illumination light path and OCT mirror [11]. Through the focusing lens and CCD camera, the images of the two parallel lights are collected, individually. The focusing system can adjust the range of the camera-receiving surface, so as to obtain the appropriate fundus retinal images. The common objective lens and the structure in front of the objective lens of the imaging system are mainly designed by using principle of Kepler telescope system, and the parallax principle of binocular vision is used in the two-way camera shooting after sharing the objective lens.

Fig. 2

Simulation diagram of imaging path

Visual matching model

The physical world is imaged on the photo sensor through the camera model, and is stored on the storage device after analogue–digital conversion. In this process, it involves the transformation of four coordinate systems, including the world coordinate system (XW, YW, ZW), the camera coordinate system (XC, YC, ZC) and the image coordinate system (XP, YP, ZP).

Linear model

To transform a point P in space into a point P′ in the image, it is necessary to master the physical meaning and mathematical transformation relationship of the four coordinate systems. Without considering the distortion, that is, under the ideal perspective condition, the principle of transformation of object point into image point is shown in Figure 3. P is located in the real space, and its position relationship can be accurately expressed only by determining the reference frame in the 3D scene [12]. The coordinate system with OW is called as the origin, and XW, YW and ZW as the axes of the free reference coordinate system, where the results obtained after the transformation of the following three coordinate systems are also based on this coordinate system. Under the reference coordinate system, the object point to the image point has to undergo the transformation of three coordinate systems, which are the camera coordinate whose origin is the optical centre of the camera; system that paralleled to the camera coordinate axis, where the image coordinate system on the ZC axis of origin; and the pixel coordinate system in the same plane as the image coordinate system. The unit of camera coordinate system and image coordinate system is the actual length unit, and the pixel coordinate system displayed by the computer needs to be transformed by the actual length of each pixel point to obtain the digital pixel coordinate [13].

Fig. 3

Transformation of coordinate system

Point P in space is a point of the actual object, and the difference of coordinates is only relative to different reference systems, which shows that reference is very important when determining the position of spatial points. The world coordinate system is transformed into the camera coordinate system through transformation of rotation and translation, and the mathematical expression of coordinate transformation in different reference frames is shown in Eq. (1). [XCYCZC]=R[XwYwZw]+T=[R11R12R13R21R22R23R31R32R33][XwYwZw]+[TxTyTz] \left[ {\matrix{ {{X_C}} \cr {{Y_C}} \cr {{Z_C}} \cr } } \right] = R\left[ {\matrix{ {{X_w}} \cr {{Y_w}} \cr {{Z_w}} \cr } } \right] + T = \left[ {\matrix{ {{R_{11}}{R_{12}}{R_{13}}} \cr {{R_{21}}{R_{22}}{R_{23}}} \cr {{R_{31}}{R_{32}}{R_{33}}} \cr } } \right]\left[ {\matrix{ {{X_w}} \cr {{Y_w}} \cr {{Z_w}} \cr } } \right] + \left[ {\matrix{ {Tx} \cr {Ty} \cr {Tz} \cr } } \right] where R is the rotation matrix, representing the trigonometric transformation combination formula of rotation angle around X, Y and Z axes; and T is the horizontal shift vector, which can be transformed into a homogeneous form, as shown in Eq. (2). [XcYcZc1]=[RT01][XwYwZw1] \left[ {\matrix{ {{X_c}} \cr {{Y_c}} \cr {{Z_c}} \cr 1 \cr } } \right] = \left[ {\matrix{ {RT} \cr {01} \cr } } \right]\left[ {\matrix{ {{X_w}} \cr {{Y_w}} \cr {{Z_w}} \cr 1 \cr } } \right]

When solving the 3D coordinates of an object, as long as the reference object is selected and the position is fixed, a camera corresponds to a certain external parameter. P′ is also at the same point of image coordinates and pixel coordinates, of which the transformation mathematical expression is shown in Eq. (3). Xp=XIdx+X0Yp=YIdy+Y0 \matrix{ {{X_p} = {{{X_I}} \over {dx}} + {X_0}} \cr {{Y_p} = {{{Y_I}} \over {dy}} + {Y_0}} \cr }

After the image points are collected by the camera, they are projected onto the imaging surface, and the computer stores them in the form of a two-dimensional array, where they are referred to as a pixel point; and each value of the two-dimensional array is called a pixel, which can be converted into matrix form as shown in Eq. (4). [XpYp1]=[1dx0X001dyY0001][XIYI1] \left[ {\matrix{ {{X_p}} \cr {{Y_p}} \cr 1 \cr } } \right] = \left[ {\matrix{ {{1 \over {{d_x}}}0{X_0}} \cr {0{1 \over {{d_y}}}{Y_0}} \cr {001} \cr } } \right]\left[ {\matrix{ {{X_I}} \cr {{Y_I}} \cr 1 \cr } } \right]

At this point, the conversion relationship between the four different coordinate systems can be connected through the camera coordinate system and the image coordinate system, and the conversion between these two coordinate systems is obtained under the ideal pinhole camera imaging model.

The P point of the object is imaged at the P′ point through the lens, and the actual image is on the plane behind the lens. However, for convenience, we move the focal plane to the front of the lens’ optical centre, and the distance between the camera optical centre and the origin of the image coordinate system is the focal length value, f, of the lens. The conversion formula of the two coordinate systems can be obtained by the principle of geometric knowledge, as shown in Eq. (5). XLXc=fZcYIYc=fZc \matrix{ {{{{X_L}} \over {{X_c}}} = {f \over {{Z_c}}}} \cr {{{{Y_I}} \over {{Y_c}}} = {f \over {{Z_c}}}} \cr } which can be expressed in form of matrix as follows: Zc[XIYI1]=[f0000f000010][XcYcZc1] {Z_c}\left[ {\matrix{ {{X_I}} \cr {{Y_I}} \cr 1 \cr } } \right] = \left[ {\matrix{ {f000} \cr {0f00} \cr {0010} \cr } } \right]\left[ {\matrix{ {{X_c}} \cr {{Y_c}} \cr {{Z_c}} \cr 1 \cr } } \right]

Combining the above formulas, the conversion formula of point P in spatial coordinates projected to P′ in pixel coordinates is as follows: ZC[XpYp1]=[1dx0X001dyY0001][f0000f000010][RT01][XwYwZw1]=[fx0X000fy0Y000010][RT01][XwYwZw1]=M[XwYwZw1] {Z_C}\left[ {\matrix{ {{X_p}} \cr {{Y_p}} \cr 1 \cr } } \right] = \left[ {\matrix{ {{1 \over {{d_x}}}0{X_0}} \cr {0{1 \over {{d_y}}}{Y_0}} \cr {001} \cr } } \right]\left[ {\matrix{ {f000} \cr {0f00} \cr {0010} \cr } } \right]\left[ {\matrix{ {RT} \cr {01} \cr } } \right]\left[ {\matrix{ {{X_w}} \cr {{Y_w}} \cr {{Z_w}} \cr 1 \cr } } \right] = \left[ {\matrix{ {{f_x}0{X_0}0} \cr {0{f_y}0{Y_0}0} \cr {0010} \cr } } \right]\left[ {\matrix{ {RT} \cr {01} \cr } } \right]\left[ {\matrix{ {{X_w}} \cr {{Y_w}} \cr {{Z_w}} \cr 1 \cr } } \right] = M\left[ {\matrix{ {{X_w}} \cr {{Y_w}} \cr {{Z_w}} \cr 1 \cr } } \right] where M is a 3 × 4 matrix, which represents the internal and external parameters of the camera. To determine their specific values, it is necessary to calibrate the camera, and the accuracy of its values will affect the accuracy of numerical solution of 3D coordinates.

Non-linear model

The above situation is a pinhole camera imaging model, that is, a linear model, where the imaging coordinates are linearly related to the coordinates of spatial physical points. However, due to the lens process and the installation deviation between sensors and lenses, the lens system will be distorted, which often leads to the non-linear correspondence between points and a distortion in the imaging system [14, 15]. For this distortion problem, in this paper, a distortion coefficient is introduced to correct it, which is transformed into a linear model.

The lens system usually has two forms of distortion, as shown in Figure 4, which are radial variation along the circle diameter direction and tangential variation along the circle circumference direction. When the ideal imaging point is located at point A because of the distortion, the position of the ideal point is shifted, and the actual imaging point is point B. In Figure 4, dr is radial change and dt is tangential change.

Fig. 4

Schematic diagram of non-linear model transformation

The radial distortion of the image not only affects the visual effect of the image but also increases the difficulty of calculation in 3D coordinates after feature point matching [16]. In reality, most lenses are circularly symmetric, and the distortion changes with the distance from the centre of the circle. Generally, Taylor series near the distortion centre is used to expand the first three terms [17]. Thus, the formula of distortion of a point in the radial direction is shown in Eqs (8) and (9). XP'=XP+(XPXo)(K1R2+K2R4+K3R6) X_P^\prime = {X_P} + \left( {{X_P} - {X_o}} \right)\left( {{K_1}{R^2} + {K_2}{R^4} + {K_3}{R^6}} \right) YP'=YP+(YPYo)(K1R2+K2R4+K3R6) Y_P^\prime = {Y_P} + \left( {{Y_P} - {Y_o}} \right)\left( {{K_1}{R^2} + {K_2}{R^4} + {K_3}{R^6}} \right)

Eq. (10) shows the change of coordinate position before and after distortion, where (Xo, Yo) is the centre point of distortion. K1, K2, K3 are the distortion coefficients. Among them, R represents the Euclidean distance between the point before distortion and the centre of distortion. Tangential distortion is caused by the misalignment between the plane and the sensor plane, which causes the image to have errors in the tangential direction. R=(XPXo)2+(YPYo)2XI'=XI+[2P1XI+P2(R2+2XI2)]YI'=YI+[2P1YI+P2(R2+2YI2)] \matrix{ {\;\;R = \sqrt {{{\left( {{X_P} - {X_o}} \right)}^2} + {{\left( {{Y_P} - {Y_o}} \right)}^2}} } \hfill \cr {X_I^\prime = {X_I} + \left[ {2{P_1}{X_I} + {P_2}\left( {{R^2} + 2X_I^2} \right)} \right]} \hfill \cr {\;Y_I^\prime = {Y_I} + \left[ {2{P_1}{Y_I} + {P_2}\left( {{R^2} + 2Y_I^2} \right)} \right]} \hfill \cr } where P1 and P2 indicate the tangential distortion coefficients. To sum up, there are five distortion coefficients, which can eliminate the distortion and transform the non-linear camera model into a linear model.

Rectify coordinate point deviation

Assuming that the position of the point in the object space is determined, the coordinate information of the point in the pixel coordinate system has also been determined. From Eq. (10), it can be found that this is a linear equation set consisting of two equations, but it contains three unknowns, and the equation set has no definite solution from the knowledge of mathematical equations. Only by finding another set of equations in the same form can the accurate information be determined, that is, only when two cameras image the same object can the 3D coordinate information of space object points be determined.

The coordinates of the pixel points of the same object point in two images can be found by technical means, and the 3D coordinates of the space point can be determined by mathematical calculation after obtaining two sets of equations. However, as the calculation procedure associated with this method is complicated, we usually calculate the positional deviation of image points, that is, parallax; then the depth information can be obtained by using the similar triangle principle, and finally the above equations are substituted to obtain the 3D coordinate information [18]. It can be seen from Figure 5 that any point in space will be at the intersection of two straight lines, and that this point is unique. Assuming that the two cameras are of the same model, the optical centre distance is T and the focal length is f; so, the depth information Z can be calculated according to the similar triangle principle, and the calculation formula is shown in Eq. (11). T(XLXR)Zf=TZ {{T - \left( {{X_L} - {X_R}} \right)} \over {Z - f}} = {T \over Z}

The depth information, namely the vertical distance Z, can be obtained from Eq. (12): Z=fTXLXR=fTd Z = {{fT} \over {{X_L} - {X_R}}} = {{fT} \over d} where XL − XR is the deviation value of the same point of an object relative to the coordinate origin in different images, which is regarded as parallax d. Owing to the existence of positive and negative values of the coordinate axis, combined with the different positions of the spatial points of the three objects in the figure, depth information of the spatial points can be ascertained based on Eq. (12), and it can also be inferred that there is an inverse proportional relationship between depth information Z and parallax d.

Fig. 5

Imaging model of deviation point

Establishment of inherent constraints of images

Feature matching is the most important link in the process of 3D reconstruction, which directly determines the accuracy of 3D coordinate points. The coordinates of two correctly matched pixel points can directly calculate d, determine the distance information of Z-axis and then calculate the X and Y values [19, 20]. In this paper, the advantages of global matching and local matching are integrated, and the window is selected to calculate the cost of feature matching. The position and size of the window are determined according to the numerical value, then the whole pixel with the window is traversed to verify whether it meets the matching result and finally according to the learning method, the best corresponding pixel position is determined after calculation and comparison, carried out multiple times.

According to the principle of visual imaging mentioned above, the calculation model of images under atmospheric scattering can be obtained as the following: I(x)=t(x)J(x)+(1t(x))A I(x) = t(x)J(x) + (1 - t(x))A

Figure 6 is an inherently constrained imaging model. From a geometric point of view, in the RGB colour space model, the vectors J(x), I(x) and A are in a state where the three points are coplanar and their endpoints are collinear. The value of the scene transmittance t(x) determines the ratio of two line segments. When t(x) is completely equal to 1, it represents a clear image with no constraint at all. For any single pixel, according to Eq. (13), all the pixels of a single foggy image are observed through point A in RGB colour space. If the position of this pixel is constrained, the distance between the grey value I(x) corresponding to this pixel X is close to the atmospheric light value A. If there is no boundary constraint at all, I(x) and J(x) coincide in the same position. The formation of inherent constraints is essentially a process of linear extrapolation where the undegraded unconstrained image can be restored by linear extrapolation along the direction from A to I(x); and the more outward the extrapolation, the more obvious the effect will be.

Fig. 6

Imaging model under inherent constraints

According to the idea of linear extrapolation, Eq. (13) is deformed, and the following formula can be derived: 1t(x)=J(x)AI(x)A {1 \over {t(x)}} = {{\parallel J(x) - A\parallel } \over {\parallel I(x) - A\parallel }}

All pixel values of each image have boundary constraints, assuming that the boundary value of each colour channel in the image J(x) is C0, C1, that is: C0J(x)C1,xΩ {C_0} \le J(x) \le {C_1},\quad \forall x \in \Omega

As shown in Figure 7, for any pixel in the image x, the linear extrapolation of vector J(x) must follow the formed physical model, and the extrapolation process must be located at the radiation cube within the minimum and maximum boundary points C0 and C1.

Fig. 7

Schematic diagram of image inherent constraint radiation

Error analysis
Different test distances

The errors that affect the visual matching effect include the measurement of the distance between optical centres, corner detection position error and feature point detection and matching in camera construction. This paper mainly studies the influence of the distance between cameras and the measured depth information on Z-axis coordinate points, the distances between cameras are set as 50–70 mm and the depth information of the same position is measured every 5 mm. Next, the distance between cameras is fixed, and the errors between the measured distance and the real distance are compared at every 100 mm between 100 mm and 5000 mm from the camera. The partial results are shown in Table 1:

Analysis of visual matching error under different test distances

Serial number Distance (mm) Actual coordinates (mm) Test coordinates (mm) Error (%)

1 50 600 607.95 1.59
2 55 600 608 1.6
3 60 600 607.55 1.51
4 65 600 606 1.3
5 70 600 606.9 1.34

It can be concluded from the table that there is an optimal measuring distance between two cameras, and when the camera distance is about 65 mm, the measuring accuracy is the highest. At the same time, it can also be found that the depth and parallax are inversely proportional and non-linear. When the parallax is very small, the small change can causes a large change in the measurement distance, while when the parallax is very large, the change of small parallax will not cause the change of measurement distance.

Corner test

Under the constraint of inherent conditions, the error of corner detection position mainly lies in the error of pixel position during detection. Since the corner has a certain size, and the number of pixels in the digital image is large and the actual length is small, it may not be the central position when the position is determined, and the actual length corresponds to an offset. Therefore, the error of the calculated internal and external parameters can be effectively analysed.

Corner test is implemented by matching algorithm. First, the 9 × 9 window is used to select the pixel value of the left image, and the pixel value of the right image is calculated by the same calculation method, so as to further compare the sizes of two pixel values. First, when the difference between the two values is less than a certain threshold, it is considered that the initial matching point has been found, the pixel coordinates of the centre point are recorded and all pixel points are matched in turn. Then, we calculate the sum of all pixel values of the matching points of the two images, and take the sum and difference of the pixels of the two images as a cost function. Finally, according to the value of the cost function, we determine whether the next re-matching calculation is needed and if the matching points need to be polished, re-select the starting position of the window according to the form of genetic mutation learning and then determine the centre point to start the next cycle until the matching cost function meets the requirements to finish the calculation and output the best matching result.

The selection of the initial window starts with the origin of the left image, and the matching of the right image also starts with the origin, except that each movement of the left image corresponds to six window movements of the right image. That is, every time the left image moves, the right image moves 2, 6 and 12 pixels in the right and down directions, respectively. Afterwards, the sum of pixel values in the window is calculated, and the threshold value is compared to determine the centre position of the matching start. Similarly, after the right image is moved for six times, the left image is moved for the next round, and the moving mode is to move 2 pixels to the right and down each time.

From Figure 8, we can see that the detection results based on ordinary visual matching have a high degree of confusion and many error points are detected. In contrast, the detection results after considering inherent constraints are very accurate, the accuracy rate is increased from 60.1% to 85.7% and almost all the detected points are corner points. Further statistics are computed based on the test results indicated in Table 2, where the experimental results are analysed and explained in detail using specific data.

Fig. 8

Corner test result chart (left: general visual matching; right: under inherent constraints)

Corner extraction results

Evaluating indicator Test method

Correct number of angular points Wrong number of angular points Undetected number of corners Accuracy (%)

General visual matching 52 26 8 60.1
Considering inherent constraints 54 3 6 85.7

Visual matching under inherent constraints filters out most pixels that are not corners through pre-screening, thus reducing the probability of errors and effectively controlling the number of detected wrong corners. The original model can also detect 54 correct corners, and the performance of the two algorithms is similar in detecting correct corners. However, in terms of the detected wrong corners, by considering the inherent constraints, only three wrong corners were detected, while the original model detected 26 wrong corners, which was about nine times that of the improved algorithm.

Conclusion

Making research on the precise error control of visual matching in geometric calculation can realise high-precision measurement. In order to solve the shortcomings of the visual matching model, such as large error of camera internal and external parameters, low-matching accuracy of feature points and long running time, etc. in this paper, the inherent constraints of images are introduced, and a non-linear visual matching model considering the constraints is established. Test results show that under the inherent constraints, there is an optimal measuring distance between two cameras, and the relationship between depth and parallax is inversely proportional, but non-linear. In visual matching under inherent constraints, most pixels that are not corners are screened out, thus reducing the probability of error, and its matching accuracy rate is increased from 60.1% to 85.7%.

Fig. 1

Cross-sectional view of visual imaging system
Cross-sectional view of visual imaging system

Fig. 2

Simulation diagram of imaging path
Simulation diagram of imaging path

Fig. 3

Transformation of coordinate system
Transformation of coordinate system

Fig. 4

Schematic diagram of non-linear model transformation
Schematic diagram of non-linear model transformation

Fig. 5

Imaging model of deviation point
Imaging model of deviation point

Fig. 6

Imaging model under inherent constraints
Imaging model under inherent constraints

Fig. 7

Schematic diagram of image inherent constraint radiation
Schematic diagram of image inherent constraint radiation

Fig. 8

Corner test result chart (left: general visual matching; right: under inherent constraints)
Corner test result chart (left: general visual matching; right: under inherent constraints)

Analysis of visual matching error under different test distances

Serial number Distance (mm) Actual coordinates (mm) Test coordinates (mm) Error (%)

1 50 600 607.95 1.59
2 55 600 608 1.6
3 60 600 607.55 1.51
4 65 600 606 1.3
5 70 600 606.9 1.34

Corner extraction results

Evaluating indicator Test method

Correct number of angular points Wrong number of angular points Undetected number of corners Accuracy (%)

General visual matching 52 26 8 60.1
Considering inherent constraints 54 3 6 85.7

Ma Songde, Zhang Zhengyou. Computer Vision-Fundamentals of Computing Theory and Algorithms (First Edition) [M]. Beijing: Science Press, 1998. (in Chinese). MaSongde ZhangZhengyou Computer Vision-Fundamentals of Computing Theory and Algorithms (First Edition) [M]. Beijing Science Press 1998 (in Chinese). Search in Google Scholar

Wheatstone C.. Contributions to the physiology of vision.—Part the first. on some remark-able, and hitherto unobserved, phenomena of binocular vision [J]. Philosophical Transactions of the Royal Society of London, 1994, 49(128). WheatstoneC. Contributions to the physiology of vision.—Part the first. on some remark-able, and hitherto unobserved, phenomena of binocular vision [J]. Philosophical Transactions of the Royal Society of London 1994 49 128 Search in Google Scholar

Dai Zongxian. Research on 3D reconstruction and measurement technology based on binocular vision [D]. Chongqing University, 2014. (in Chinese). DaiZongxian Research on 3D reconstruction and measurement technology based on binocular vision [D]. Chongqing University 2014 (in Chinese). Search in Google Scholar

Xu Zhou Peng, Yang Chaolin. On-line detection method of tiny defects on the surface of sheet metal strip based on photometric stereology [J]. Journal of Mechanical Engineering. 2013, 49(04). (in Chinese). XuZhou Peng YangChaolin On-line detection method of tiny defects on the surface of sheet metal strip based on photometric stereology [J]. Journal of Mechanical Engineering 2013 49 04 (in Chinese). 10.3901/JME.2013.04.025 Search in Google Scholar

He Qing. Research on 3D reconstruction technology based on binocular stereo vision [D]. Huazhong University of Science and Technology. (in Chinese). HeQing Research on 3D reconstruction technology based on binocular stereo vision [D]. Huazhong University of Science and Technology (in Chinese). Search in Google Scholar

Wang Peng. Research on target recognition and location based on binocular vision [D]. Yanshan University, 2020. (in Chinese). WangPeng Research on target recognition and location based on binocular vision [D]. Yanshan University 2020 (in Chinese). Search in Google Scholar

Sdiri B, Kaaniche M, Cheikh F A, et al. Efficient enhancement of stereo endoscopic images based on joint wavelet decomposition and binocular combination [J]. IEEE transactions on medical imaging, 2018. SdiriB KaanicheM CheikhF A Efficient enhancement of stereo endoscopic images based on joint wavelet decomposition and binocular combination [J]. IEEE transactions on medical imaging 2018 10.1109/TMI.2018.285380829994612 Search in Google Scholar

Cao Jing. Research and implementation of key technologies in binocular stereo vision system [D]. Xi ’an University of Science and Technology, 2019. (in Chinese). CaoJing Research and implementation of key technologies in binocular stereo vision system [D]. Xi ’an University of Science and Technology 2019 (in Chinese). Search in Google Scholar

Rosten E, Drummond T. Machine Learning for High-Speed Corner Detection. European Conference on Computer Vision. Springer-Verlag, 2006:430–443P. RostenE MachineDrummond T Learning for High-Speed Corner Detection. European Conference on Computer Vision Springer-Verlag 2006 430 443 10.1007/11744023_34 Search in Google Scholar

Xin Yang. Research on the key technologies of computer-generated holographic 3D display [D]. Soochow University, 2018. (in Chinese). XinYang Research on the key technologies of computer-generated holographic 3D display [D]. Soochow University 2018 (in Chinese). Search in Google Scholar

Yu Daoyin. Fundamentals of Engineering Optics [M]. Beijing: China Machinery Industry Press, 2011. (in Chinese). YuDaoyin Fundamentals of Engineering Optics [M]. Beijing China Machinery Industry Press 2011 (in Chinese). Search in Google Scholar

Cassisa C.. Local vs global energy minimization methods: application to stereo matching [A]. In: IEEE International Conference on Progress in Informatics and Computing [C],2010: 678–683. CassisaC. Local vs global energy minimization methods: application to stereo matching [A]. In: IEEE International Conference on Progress in Informatics and Computing [C], 2010 678 683 Search in Google Scholar

Kuzmin A., Mikushin D., Lempitsky V. End-to-End learning of cost-volume aggregation for real-time dense stereo [A]. In: IEEE International Workshop on Machine Learning for Signal Processing [C], 2017: 1–6. KuzminA. MikushinD. LempitskyV End-to-End learning of cost-volume aggregation for real-time dense stereo [A] In: IEEE International Workshop on Machine Learning for Signal Processing [C], 2017 1 6 10.1109/MLSP.2017.8168183 Search in Google Scholar

Lv eryang. Research on a design method of relay framing optical system [D]. Tianjin University, 2012. (in Chinese). Lveryang Research on a design method of relay framing optical system [D]. Tianjin University 2012 (in Chinese). Search in Google Scholar

Lowe D G. Distinctive Image Features from Scale-Invariant Keypoints. International Journal of Computer Vision, 2004, 60(2):91–110P. LoweD G Distinctive Image Features from Scale-Invariant Keypoints International Journal of Computer Vision 2004 60 2 91 110 10.1023/B:VISI.0000029664.99615.94 Search in Google Scholar

Ke Y, Sukthankar R. PCA-SIFT: a more distinctive representation for local image descriptors//IEEE Computer Society Conference on Computer Vision and Pattern Recognition. IEEE Computer Society, 2004:28–43P. KeY SukthankarR PCA-SIFT: a more distinctive representation for local image descriptors IEEE Computer Society Conference on Computer Vision and Pattern Recognition. IEEE Computer Society 2004 28 43 Search in Google Scholar

Shi Erying, Zhu Jiaqun, Yang Changchun. Image forgery detection algorithm based on nearest neighbor search coupled with nearest neighbor loss clustering. Packaging Engineering, 2018(5):185–190. (in Chinese). ShiErying ZhuJiaqun YangChangchun Image forgery detection algorithm based on nearest neighbor search coupled with nearest neighbor loss clustering Packaging Engineering 2018 5 185 190 (in Chinese). Search in Google Scholar

Suo Chunbao, Yang Dongqing, Liu Yunpeng. Comparison of SIFT, SURF, BRISK, ORB and FREAK algorithms from various angles [J]. Beijing Surveying and Mapping, 2014, 04(06). (in Chinese). SuoChunbao YangDongqing LiuYunpeng Comparison of SIFT, SURF, BRISK, ORB and FREAK algorithms from various angles [J]. Beijing Surveying and Mapping 2014 04 06 (in Chinese). Search in Google Scholar

Ghani A S A, Isa N A M. Enhancement of low quality underwater image through integrated global and local contrast correction [J]. Applied Soft Computing, 2015, 37: 332–344. GhaniA S A IsaN A M Enhancement of low quality underwater image through integrated global and local contrast correction [J]. Applied Soft Computing 2015 37 332 344 10.1016/j.asoc.2015.08.033 Search in Google Scholar

Ghani A S A, Isa N A M. Automatic system for improving underwater image contrast and color through recursive adaptive histogram modification [J]. Computers and electronics in agriculture, 2017, 141: 181–195. GhaniA S A IsaN A M Automatic system for improving underwater image contrast and color through recursive adaptive histogram modification [J]. Computers and electronics in agriculture 2017 141 181 195 10.1016/j.compag.2017.07.021 Search in Google Scholar

Articles recommandés par Trend MD

Planifiez votre conférence à distance avec Sciendo