Fusion method and system of multi-depth information
Through nonlinear conversion and optical center conversion technology, the problems of traditional deep information fusion median expansion and low precision are solved, and high-precision multi-depth information fusion and de-mask effects are achieved.
Patent Information
- Application Number
- CN202110413690.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2020-04-16
- Filing Date
- 2021-04-16
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2041-04-16
AI Technical Summary
In stereoscopic display technology, the traditional deep information fusion method causes a significant expansion of the number of bits through linear conversion, making it impossible to achieve high precision deep information fusion, and the de-masking effect is poor.
The nonlinear conversion method is used to convert and fusion of depth information, combined with optical center conversion technology, reduce the number of bits and improve the accuracy.
It effectively reduces the number of bits expansion, improves the accuracy of depth information, and makes full use of the highest precision depth information in the de-mask program, significantly improving the accuracy of the fusion result.
Smart Images

Figure CN113537276B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a fusion method and a fusion system, and in particular to a fusion method and a fusion system for multi-depth information. Background Art
[0002] With the rapid development of display technology, a stereoscopic display technology has been developed. In stereoscopic display technology, depth information of multiple different baselines / depth disparities can be fused to obtain various advantages. In order to make the depth information of different depth disparities be on the same basis, a normalization operation must be performed. Generally speaking, the normalization operation is to multiply the depth disparity by the same value in a linear conversion manner. This will lead to a significant expansion of the number of bits.
[0003] In addition, the fusion of multiple depth information can achieve the effect of de-masking, but in the traditional de-masking process, the length of the baseline is not long, and it is impossible to obtain high-precision depth information. Summary of the invention
[0004] The present invention relates to a fusion method and fusion system for multi-depth information, which performs conversion by nonlinear conversion so that the number of bits will not be greatly expanded. In addition, when performing a de-masking process, an optical center conversion technology is used to fully utilize the depth information with the highest accuracy, so that the fusion result can greatly improve the accuracy.
[0005] According to one aspect of the present invention, a method for fusing multiple depth information is proposed. The method for fusing multiple depth information comprises the following steps. A first depth information is obtained by using a first image acquisition module. A second depth information is obtained by using a second image acquisition module. The first depth information is converted into a first conversion information by a nonlinear conversion and the second depth information is processed according to the first conversion information to generate the converted information; or after fusing the first depth information and the second depth information into a fused information, the fused information is converted into the converted information by the nonlinear conversion.
[0006] According to another aspect of the present invention, a fusion system of multiple depth information is proposed. The fusion system of multiple depth information includes a first image acquisition module, a second image acquisition module, a depth generation unit and a fusion unit. The depth generation unit is used to obtain a first depth information based on a first image acquisition module and obtain a second depth information based on a second image acquisition module. The fusion unit transforms the first depth information into a first transformation information through a nonlinear transformation, and processes the second depth information based on the first transformation information to generate the transformed information, or after fusing the first depth information and the second depth information into a fused information, transforms the fused information into the transformed information through the nonlinear transformation.
[0007] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments, but is not intended to limit the present invention. BRIEF DESCRIPTION OF THE DRAWINGS
[0008] Figure 1 FIG. 4 is a schematic diagram of a system for fusing multiple depth information according to an embodiment.
[0009] Figure 2 FIG. 4 is a flowchart of a method for fusing multiple depth information according to an embodiment.
[0010] Figure 3 is a schematic diagram of parallax according to an embodiment.
[0011] Figure 4 A schematic diagram illustrating a linear conversion according to an embodiment is shown.
[0012] Figure 5 A schematic diagram illustrating a nonlinear transformation according to an embodiment.
[0013] Figure 6 is a circuit diagram of a nonlinear conversion according to an embodiment.
[0014] Figure 7 is a circuit diagram of nonlinear conversion according to another embodiment.
[0015] Figure 8 A schematic diagram illustrating nonlinear conversion according to another embodiment is shown.
[0016] Fig. 9 is a schematic diagram of a system for fusing multiple depth information according to another embodiment.
[0017] Fig.10 is a flowchart of a method for fusing multiple depth information according to another embodiment.
[0018] Wherein, the reference numerals are:
[0019] 100,100': Fusion System
[0020] 110A: first image acquisition unit
[0021] 110B: Second image acquisition unit
[0022] 110C: third image acquisition unit
[0023] 120,120': Depth generation unit
[0024] 121: First Depth Engine
[0025] 122: Second Depth Engine
[0026] 123: The Third Depth Engine
[0027] 140: Optical center conversion unit
[0028] 150: Fusion Unit
[0029] 151:Multiplier
[0030] 152: Subtractor
[0031] 153: Divider
[0032] 154: Adder
[0033] Bab: First Baseline
[0034] Bac: Second Baseline
[0035] BBC: Third Baseline
[0036] D1,D1',D1",D1': data
[0037] Dab: First Depth Parallax
[0038] Dac: Second Depth Parallax
[0039] DM1: First depth information
[0040] DM2: Second depth information
[0041] DM3: Third Depth Information
[0042] P1, P2: Turning Point
[0043] r: ratio
[0044] S110,S120,S140,S150,S150': Steps
[0045] thd,thd': critical value DETAILED DESCRIPTION
[0046] The structural principle and working principle of the present invention are described in detail below in conjunction with the accompanying drawings:
[0047] Please refer to Figure 1, which is a schematic diagram of a fusion system 100 of multiple depth information according to an embodiment. The fusion system 100 includes a first image acquisition module (not shown in the figure) composed of a first image acquisition unit 110A and a second image acquisition unit 110B, a second image acquisition module (not shown in the figure) composed of the second image acquisition unit 110B and a third image acquisition unit 110C, a depth generation unit 120 and a fusion unit 150. The first image acquisition unit 110A, the second image acquisition unit 110B, and the third image acquisition unit 110C are used to acquire color images, grayscale images, or infrared images, such as color cameras, black and white cameras, or infrared cameras. The depth generation unit 120, for example, includes a first depth engine 121 and a second depth engine 122. In another embodiment, the depth generation unit 120 may include only one depth engine. The fusion unit 150 is used to fuse multiple depth information. The depth generation unit 120 and / or the fusion unit 150 is, for example, a circuit, a chip, a cloud computing center, a program code, a circuit board, or a storage device for storing program code. In the process of fusion of depth information, the depth information must be transformed. However, the conversion process may cause a significant expansion of the number of bits. The fusion system 100 of this embodiment can be transformed through nonlinear conversion so that the number of expanded bits can be reduced and the amount of stored data can be reduced. The following is a flowchart to describe in detail the operation of the fusion system 100 of this embodiment.
[0048] Please refer to Figure 2 , which is a flow chart of a method for fusing multiple depth information according to an embodiment. In step S110, according to the first image acquisition unit 110A and the second image acquisition unit 110B, the first depth engine 121 of the depth generation unit 120 obtains a first depth information DM1. A first baseline Bab is located between the first image acquisition unit 110A and the second image acquisition unit 110B. The first baseline Bab is, for example, 3 cm. Please refer to Figure 3 , which is a schematic diagram of disparity according to an embodiment. The first depth information DM1 has a first depth disparity Dab. The first depth information DM1 referred to in the present invention refers to the entire picture.
[0049] Next, in step S120, the second depth engine 122 of the depth generating unit 120 obtains a second depth information DM2 according to the first image acquisition unit 110A and the third image acquisition unit 110C. A second baseline Bac is located between the first image acquisition unit 110A and the third image acquisition unit 110C. The second baseline Bac is, for example, 12 cm. The second depth information DM2 has a second depth disparity Dac. The second baseline Bac is greater than the first baseline Bab, and the ratio of the second baseline Bac to the first baseline Bab is, for example, 4. The second depth information DM2 referred to in the present invention refers to the entire picture.
[0050] The above-mentioned step S110 and step S120 can be executed simultaneously or in a different order, and the present invention is not limited to the execution order.
[0051] Then, in step S150, the fusion unit 150 performs fusion according to the first depth information DM1 and the second depth information DM2. In step S150, the fusion unit 150 transforms the first depth information DM1 into transformed information through a nonlinear transformation. In the process of fusion of depth information, the depth information must be transformed. However, the traditional conversion process may cause a significant expansion of the number of bits. After the nonlinear transformation of the present invention is used to transform the transformed information, the number of expanded bits can be reduced and the amount of stored data can be reduced. The transformed information referred to in the present invention refers to the entire picture.
[0052] Conventionally, the first depth information DM1 is converted by linear transformation. For example, the following equation (1) is a linear transformation method. The second baseline Bac has a ratio r with the first baseline Bab.
[0053] D1'=D1*r…………………………………………(1)
[0054] Please refer to Figure 4 , which illustrates a schematic diagram of linear conversion according to an embodiment. The data D1 originally "0, 1, 2, 3" is converted through linear conversion to data D1' such as "0, 4, 8, 12". In order to store the data D1' such as "0, 4, 8, 12", an additional 2 bits are required, resulting in a significant expansion of the number of bits. Between the values 4, 8, 12, there are no values 5, 6, 7, 9, 10, 11, etc., which wastes a lot of data. When the data D1 is linearly converted according to the above formula (1), the converted data D1' jumps its value with a large amplitude, thereby resulting in a significant expansion of the number of bits. The data D1 and data D1' referred to in the present invention refer to the image value of a pixel.
[0055] In order to avoid a significant expansion of the number of bits, this embodiment performs the conversion by nonlinear conversion. Figure 5, which illustrates a schematic diagram of nonlinear conversion according to an embodiment. In order to reduce the expansion of the number of bits, the data D1' is nonlinearly converted according to the following formula (2) to obtain the converted data D1". The data D1" referred to in the present invention refers to the image value of a pixel.
[0056] D1”=(D1’>thd)?thd+(D1’-thd) / r:D1’………………(2)
[0057] That is, if the data D1' is higher than a threshold value thd, it is converted at a conversion ratio; if the data D1' is not higher than the threshold value thd, it remains as data D1'. The threshold value thd is, for example, the maximum value of the data of the second depth information DM2 having the second depth disparity Dac. The conversion ratio is the ratio r of the second baseline Bac to the first baseline Bab (shown in Figure 7 ), that is, 1 / r. Figure 5 As shown, the data D1" after nonlinear conversion jumps its value with a smaller amplitude. For example, in the operation of the above formula (2), the critical value thd is, for example, 4, and the ratio r is, for example, 4. The above-mentioned data D1' of "0" is not higher than the critical value thd, so the original data D1' is still used as data D1", that is, "0". The above-mentioned data D1' of "4" is not higher than the critical value thd, so the original data D1' is still used as data D1", that is, "4". The data D1' of "8" is higher than the critical value thd, so through the calculation of "critical value thd + (data D1'-critical value thd) / ratio r", the data D1" of "5 (i.e. 4 + (8-4) / 4 = 5)" is obtained. The data D1' of "12" is higher than the critical value thd, so through the calculation of "critical value thd + (data D1'-critical value thd) / ratio r", the data D1" of "6 (i.e. 4 + (12-4) / 4 = 6)" is obtained. In other words, the data D1' of "0, 4, 8, 12" can be converted into the data D1" of "0, 4, 5, 6" through the above formula (2). The data D1" after nonlinear conversion jumps its value with a smaller amplitude.
[0058] The combination of the above formula (1) and the above formula (2) can obtain the following formula (3):
[0059] D1”=(D1*r>thd)?thd+(D1*r-thd) / r:D1*r…………(3)
[0060] Please refer to Figure 6 , which is a circuit diagram of nonlinear conversion according to an embodiment. Figure 6The circuit diagram is used to realize the nonlinear conversion of the above formula (3). Data D1 and the ratio r are input to a multiplier 151 for multiplication. The output result of the multiplier 151 and the critical value thd are input to a subtractor 152 for subtraction. Then, the output result of the subtractor 152 and the ratio r are input to a divider 153 for division. Finally, the output result of the divider 153 and the critical value thd are input to an adder 154 for addition to obtain data D1".
[0061] In another embodiment, the above formula (3) can be simplified to the following formula (4):
[0062] D1"=(cond)?D1+thd':D1*r
[0063] thd'=thd-thd / r
[0064] cond, for example, is D1*r>thd……………………………………(4)
[0065] Please refer to Figure 7 , which is a circuit diagram of nonlinear conversion according to another embodiment. Figure 7 The circuit diagram is used to implement the operation of the above formula (4). The data D1 and the critical value thd' are input to the adder 154 for addition operation to obtain the data D1". The critical value thd' can be pre-stored in the memory in the form of parameters without the need for operation. Generally speaking, the circuit structure of the adder 154 is quite simple and can be implemented without too many logic elements. The circuit structure of the divider 153 is the most complex and requires quite a lot of logic elements to implement. The designer can use Figure 7 The design can reduce circuit complexity, computational delay and system load. For example, in the operation of formula (4), the critical value thd is 4, and the ratio r is 4. The data D1 such as "0, 1" does not need to be calculated by "data D1+critical value thd'". The data D1 of "2" can be calculated by "data D1+critical value thd'" to obtain data D1 of "5 (i.e. 2+3=5)". The data D1 of "3" can be calculated by "data D1+critical value thd'" to obtain data D1 of "6 (i.e. 3+3=6)". In other words, the data D1 of "2, 3" can be obtained by simple addition operation to obtain data D1 of "5, 6".
[0066] In another embodiment, the nonlinear transformation in step S150 can also be implemented by a bit right shift operation. In order to reduce the expansion of the bit number, the data D1' can be nonlinearly transformed according to the following formula (5).
[0067] D1"=(cond)?(D1>>s)+thd':D1'
[0068] cond is, for example, D1'>thd…………………………………………(5)
[0069] In the above formula (5), s is, for example, 1. That is, the bit right shift operation in the above formula (5) is a 1-bit right shift operation. For example, the data D1' originally such as "8, 12, 16, 32" will be converted into data D1" such as "4, 6, 8, 16" through linear conversion. After the 1-bit right shift operation, the value of the data D1" can be effectively reduced, and the significant expansion of the number of bits can be effectively avoided.
[0070] Please refer to Figure 8 , which illustrates a schematic diagram of nonlinear conversion according to another embodiment. In some embodiments, when three depth information need to be fused, the expansion of the number of bits will be more serious. The data D1' can be used Figure 8 The nonlinear transformation is performed in a two-stage manner. Figure 8 As shown, Figure 8 There are two turning points P1 and P2. The data D1'' after nonlinear conversion can jump its value in a smaller amplitude. In addition, the nonlinear conversion of this embodiment can also be applied to the fusion of four depth information, the fusion of five depth information, or even the fusion of an unlimited number of depth information. Just by using multiple turning points, the jump amplitude of the value can be smoothly reduced. The data D1'' referred to in the present invention refers to the image value of a pixel.
[0071] After the data D1 is converted into data D1', data D1" or data D1"', the amount of data and logic size required for the memory required for the operation can be greatly reduced. After the operation, these data D1', data D1" or data D1"' can be reversely converted back to data D1 to calculate the correct depth information.
[0072] Through the above-mentioned embodiment, the first depth information DM1 and the second depth information DM2 can be transformed by nonlinear conversion during the fusion process, so that the number of bits will not be greatly expanded.
[0073] In addition, the above-mentioned nonlinear conversion action can be performed before the fusion selection action. Alternatively, the above-mentioned nonlinear conversion action can be performed after the fusion selection action. The present invention does not limit the execution order of the nonlinear conversion action. In addition, when the nonlinear conversion action is before the fusion action, the present invention has the following different practices: first convert the first depth information DM1 into the first conversion information by nonlinear conversion, and then fuse the first conversion information and the second depth information DM2; or first convert the first depth information DM1 into the first conversion information and convert the second depth information DM2 into the second conversion information by nonlinear conversion, and then fuse the first conversion information and the second conversion information. In addition, when the nonlinear conversion action is after the fusion action, the present invention can first fuse the first depth information and the second depth information into a fused information, and then convert the fused information into the converted information by the nonlinear conversion. It is emphasized again that the present invention does not limit the execution order of the nonlinear conversion action and the fusion selection action.
[0074] Furthermore, the fusion of multiple depth information can also achieve the effect of removing occlusion. Fig. 9 , which is a schematic diagram of a multi-depth information fusion system 100' according to another embodiment. The fusion system 100' comprises a first image acquisition unit 110A, a second image acquisition unit 110B, a third image acquisition unit 110C, a depth generation unit 120, an optical center conversion unit 140 and a fusion unit 150.
[0075] Generally speaking, when performing the de-masking process, the optical centers of the two depth information must be located at the same position. Fig. 9For example, the first image acquisition unit 110A is located between the second image acquisition unit 110B and the third image acquisition unit 110C. An optical center of the first depth information DM1 is located at the first image acquisition unit 110A. An optical center of the second depth information DM2 is located at the first image acquisition unit 110A. The first baseline Bab is located between the first image acquisition unit 110A and the second image acquisition unit 110B. The first baseline Bab is, for example, 3 cm. The second baseline Bac is located between the first image acquisition unit 110A and the third image acquisition unit 110C. The second baseline Bac is, for example, 12 cm. Generally speaking, the longer the length of the baseline, the higher the accuracy of the depth information. If the de-obstruction process is performed only through the first depth information DM1 and the second depth information DM2, only the accuracy of the first baseline Bab and the second baseline Bac can be used. In this embodiment, a third baseline Bbc is located between the second image acquisition unit 110B and the third image acquisition unit 110C. The third baseline Bbc is, for example, 15 cm. This embodiment further uses the third baseline Bbc to improve the accuracy of the depth information of the de-obstruction process. The following is a flowchart to explain in detail the de-masking process performed by the fusion method of this embodiment.
[0076] Please refer to Fig.10 , which is a flow chart of a method for fusing multiple depth information according to another embodiment. Fig.10 In the fusion method, a masking process that can improve the accuracy is further implemented. In step S110, the first depth engine 121 of the depth generating unit 120' obtains the first depth information DM1 according to the first image acquiring unit 110A and the second image acquiring unit 110B.
[0077] Next, in step S120 , the second depth engine 122 of the depth generating unit 120 ′ obtains second depth information DM2 according to the first image acquiring unit 110A and the third image acquiring unit 110C.
[0078] Then, in step S130, the third depth engine 123 of the depth generating unit 120' obtains a third depth information DM3 according to the second image acquiring unit 110B and the third image acquiring unit 110C. A third baseline Bbc is located between the second image acquiring unit 110B and the third image acquiring unit 110C. The third baseline Bbc is greater than the second baseline Bac and the first baseline Bab. The third baseline Bbc is, for example, 15 cm. Since the third baseline Bbc is the longest, the third depth information DM3 has the highest accuracy.
[0079] The above steps S110, S120, and S130 can be executed simultaneously. Alternatively, the execution order of steps S110, S120, and S130 can be adjusted arbitrarily. The present invention is not limited to the execution order.
[0080] Next, in step S140, the optical center conversion unit 140 projects the third depth information DM3 to the optical center of the first depth information DM1 and the projection surface. In this step, after obtaining the spatial relationship (e.g., external parameters) between the third depth information DM3 and the first depth information DM1, the optical center conversion unit 140 projects the third depth information DM3 to the optical center of the first depth information DM1 and the projection surface according to the spatial relationship.
[0081] Then, in step S150', the fusion unit 150 performs fusion according to the first depth information DM1, the second depth information DM2 and the third depth information DM3. During step S150', the fusion unit 150 also performs a de-masking procedure. In this embodiment, when the fusion unit 150 performs the de-masking procedure, the third depth information DM3 with the highest accuracy can be fully utilized, so that the fusion result can greatly improve the accuracy.
[0082] In addition, in the present embodiment, the first image acquisition module is composed of the first image acquisition unit 110A and the second image acquisition unit 110B, and the second image acquisition module is composed of the second image acquisition unit 110B and the third image acquisition unit 110C. However, in other embodiments of the present invention, the first image acquisition module may be composed of the first image acquisition unit 110A and the second image acquisition unit 110B, and the second image acquisition module may be composed of the third image acquisition unit 110C and the fourth image acquisition unit (not shown). Since the first image acquisition module and the second image acquisition module do not share the same image acquisition unit, the optical center conversion unit 140 is required to convert the optical center before performing the nonlinear conversion of the present invention. For example, when the optical center of the first depth information is set in the first image acquisition unit and the optical center of the second depth information is set in the third image acquisition unit, the optical center conversion unit 140 is required to convert the optical centers of the two to be the same. For example, the second depth information is first projected to the optical center of the first depth information and the projection surface before performing the nonlinear conversion of the present invention.
[0083] Of course, the present invention may have many other embodiments. Without departing from the spirit and essence of the present invention, those skilled in the art may make various corresponding changes and modifications based on the present invention, but these corresponding changes and modifications should all fall within the scope of protection of the claims attached to the present invention.
Claims
1. A method for fusing multiple depth information, characterized in that: include: Obtaining first depth information using a first image acquisition module; Obtaining second depth information using a second image acquisition module; as well as After fusing the first depth information and the second depth information into fused information, converting the fused information into transformed information through nonlinear transformation; Through the nonlinear conversion, the number of bits of the converted information obtained by merging and converting the first depth information and the second depth information is equal to or smaller than the number of bits of the first depth information.
2. The method for fusing multiple depth information as claimed in claim 1, characterized in that: The non-linear transformation is performed using a transformation ratio.
3. The method for fusing multiple depth information as claimed in claim 2, characterized in that: The first image acquisition module further includes a first image acquisition unit and a second image acquisition unit, and a first baseline is located between the first image acquisition unit and the second image acquisition unit; The second image acquisition module further includes the first image acquisition unit and a third image acquisition unit, a second baseline is located between the first image acquisition unit and the third image acquisition unit; and The conversion ratio is the inverse of the ratio of the second baseline to the first baseline.
4. The method for fusing multiple depth information as claimed in claim 1, characterized in that: The nonlinear transformation converts the fused information by a bit right shift operation.
5. The method for fusing multiple depth information as claimed in claim 4, characterized in that: This bit right shift operation is a 1-bit right shift operation.
6. The method for fusing multiple depth information as claimed in claim 1, characterized in that: The first image acquisition module further includes a first image acquisition unit and a second image acquisition unit, and an optical center of the first depth information is set in the first image acquisition unit; the second image acquisition module further includes a third image acquisition unit and a fourth image acquisition unit, and an optical center of the second depth information is set in the third image acquisition unit. The fusion method further includes: The second depth information is projected onto the optical center of the first depth information and a projection surface.
7. The method for fusing multiple depth information as claimed in claim 1, characterized in that: The first image acquisition module further includes a first image acquisition unit and a second image acquisition unit, the second image acquisition module further includes the second image acquisition unit and a third image acquisition unit, the first image acquisition unit is located between the second image acquisition unit and the third image acquisition unit, an optical center of the first depth information is located in the first image acquisition unit, and an optical center of the second depth information is located in the first image acquisition unit, and the fusion method further includes: Obtaining third depth information according to the second image acquisition unit and the third image acquisition unit; Projecting the third depth information onto the optical center of the first depth information and a projection surface; and A de-occlusion process is performed according to the first depth information, the second depth information and the third depth information.
8. A multi-depth information fusion system, characterized in that: include: a first image acquisition module; a second image acquisition module; a depth generating unit, configured to obtain first depth information by using the first image acquiring module and obtain second depth information by using the second image acquiring module; as well as a fusion unit, which fuses the first depth information and the second depth information into fused information, and then transforms the fused information into transformed information through nonlinear transformation; Through the nonlinear conversion, the number of bits of the converted information obtained by merging and converting the first depth information and the second depth information is equal to or smaller than the number of bits of the first depth information.
9. The multi-depth information fusion system as claimed in claim 8, characterized in that: The non-linear transformation is performed using a transformation ratio.
10. The multi-depth information fusion system according to claim 9, characterized in that: The first image acquisition module further includes a first image acquisition unit and a second image acquisition unit, and a first baseline is located between the first image acquisition unit and the second image acquisition unit; The second image acquisition module further includes the first image acquisition unit and a third image acquisition unit, a second baseline is located between the first image acquisition unit and the third image acquisition unit; and The conversion ratio is the inverse of the ratio of the second baseline to the first baseline.
11. The multi-depth information fusion system according to claim 8, characterized in that: The nonlinear transformation converts the fused information by a bit right shift operation.
12. The multi-depth information fusion system according to claim 11, characterized in that: This bit right shift operation is a 1-bit right shift operation.
13. The multi-depth information fusion system according to claim 8, characterized in that: The first image acquisition module further includes a first image acquisition unit and a second image acquisition unit, and an optical center of the first depth information is set in the first image acquisition unit; the second image acquisition module further includes a third image acquisition unit and a fourth image acquisition unit, and an optical center of the second depth information is set in the third image acquisition unit, and the fusion system further includes: The second depth information is projected onto the optical center of the first depth information and a projection surface.
14. The multi-depth information fusion system according to claim 8, characterized in that: The first image acquisition module further includes a first image acquisition unit and a second image acquisition unit, the second image acquisition module further includes the second image acquisition unit and a third image acquisition unit, the first image acquisition unit is located between the second image acquisition unit and the third image acquisition unit, an optical center of the first depth information is located in the first image acquisition unit, an optical center of the second depth information is located in the first image acquisition unit, and a third depth information is obtained according to the second image acquisition unit and the third image acquisition unit, and the fusion system further includes: an optical center conversion unit, used for projecting the third depth information onto the optical center of the first depth information and a projection surface; The fusion unit performs a de-masking process according to the first depth information, the second depth information and the third depth information.
Citation Information
Cited By
Information fusion method and system for deep geological exploration
CN119885088A
An information fusion method and system for deep geological exploration
CN119885088B
