Image processing acceleration method and system based on multi-step phase shifting
By employing image processing methods based on fixed-point representation and pipelined architecture, combined with advanced memory and least squares method, the computational bottleneck of traditional multi-step phase-shifting method in high-speed dynamic measurement is solved, achieving efficient and low-power image processing suitable for portable or mobile measurement devices.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-03-18
- Publication Date
- 2026-07-24
AI Technical Summary
Traditional multi-step phase-shifting methods have high computational complexity when processing high-resolution, multi-frame images, making it difficult to meet the real-time and low-power requirements of high-speed dynamic measurement. They also have significant bottlenecks in high-performance computing tasks involving megapixel and multi-frame images.
An image processing method based on fixed-point representation is adopted, which combines a data caching module of first-in-first-out memory and double data rate synchronous dynamic random access memory. Through pipeline structure and least squares method, image data is acquired, format converted and synchronous pixel calculation is performed to achieve accurate acquisition of image processing results.
It improves the speed and efficiency of image processing, meets the real-time requirements of high-speed dynamic scenes, reduces system power consumption, and is suitable for portable or mobile measurement devices.
Smart Images
Figure CN122453669A_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to the field of image processing technology, and in particular to an image processing acceleration method and system based on a multi-step phase-shifting method. Background Technology
[0002] Multi-step phase-shifting method, as a high-precision fringe phase resolution technique, has been widely used in structured light 3D measurement, optical microscopy, and industrial non-destructive testing. Its core principle is to project a series of sinusoidal fringe patterns with a fixed phase difference onto the object under test, acquire images of the deformed fringe patterns modulated by the object's surface, and then use phase extraction algorithms to reconstruct the object's 3D shape or phase distribution.
[0003] Traditional multi-step phase-shifting methods rely on software implementation based on general-purpose processors. While they offer flexibility, they suffer from significant bottlenecks when processing high-resolution, multi-frame image sequences. First, the algorithm involves numerous trigonometric functions, multiplication, accumulation, and iterative operations, resulting in high computational complexity that makes it difficult to meet the real-time requirements of high-speed dynamic measurements. Second, the data transmission, storage, and processing of multi-frame images put pressure on system bandwidth and memory, limiting its application in high-speed, high-frame-rate scenarios. When dealing with high-performance computing tasks involving megapixel and multi-frame images, it is difficult to meet the dual requirements of real-time performance and low power consumption.
[0004] Therefore, existing technologies have significant bottlenecks in terms of processing speed, power consumption, and real-time performance, making it difficult to meet the requirements of high-speed acquisition and high frame rate imaging. Summary of the Invention
[0005] In view of the above problems, embodiments of this disclosure provide an image processing acceleration method and system based on a multi-step phase-shifting method.
[0006] One aspect of this disclosure provides an image processing acceleration method based on a multi-step phase-shifting method, comprising: acquiring multiple first image data of an object under test, wherein any two adjacent first image data have a fixed phase difference; determining the light intensity values of multiple pixels in each of the multiple first image data, and determining an image processing result based on the light intensity values and phase difference of the multiple pixels, wherein the image processing result characterizes the phase information and modulation information of the object under test, and the image processing result is represented by a floating-point number.
[0007] According to embodiments of this disclosure, acquiring multiple first image data of the object under test includes: sequentially acquiring multiple sinusoidal stripe patterns projected onto the surface of the object under test by a projection device in a fixed phase difference order; parsing the multiple sinusoidal stripe patterns to obtain multiple second image data, the second image data being represented by fixed-point numbers; and decoding the multiple second image data to obtain multiple first image data, the first image data being represented by floating-point numbers.
[0008] According to embodiments of this disclosure, determining an image processing result based on the light intensity values and phase differences of multiple pixels includes: determining the vertical offset and horizontal offset corresponding to the phase difference; determining the vertical projection intensity of the light intensity value of each pixel in the vertical direction based on the light intensity value and vertical offset of each pixel in different first image data; determining the horizontal projection intensity of the light intensity value of each pixel in the horizontal direction based on the light intensity value and horizontal offset of each pixel in different first image data; and determining the image processing result based on the vertical projection intensity and horizontal projection intensity of each pixel.
[0009] According to embodiments of this disclosure, determining the image processing result based on the vertical projection intensity and horizontal projection intensity of each pixel includes: calculating the arctangent function of the ratio of vertical projection intensity to horizontal projection intensity to obtain the phase information of the object under test; calculating the square root of the sum of the squares of the vertical projection intensity and the horizontal projection intensity, and multiplying it by a preset constant scaling factor to obtain the modulation information of the object under test.
[0010] According to embodiments of this disclosure, the image processing acceleration method further includes: storing vertical offset and horizontal offset in a lookup table; and determining the vertical projection intensity and horizontal projection intensity of the light intensity value of each pixel based on the light intensity value of each pixel in different first image data and the lookup table.
[0011] According to embodiments of this disclosure, the image processing acceleration method further includes: performing fixed-point format conversion on the image processing result to obtain an image processing result represented in fixed-point numbers; and sending the image processing result represented in fixed-point numbers to a host computer for result verification and visualization processing.
[0012] According to embodiments of this disclosure, the image processing acceleration method further includes: caching a plurality of first image data in a data cache module composed of a first-in-first-out memory and a double data rate synchronous dynamic random access memory; retrieving the plurality of first image data from the data cache module, and determining the light intensity values of a plurality of pixels in each of the first image data.
[0013] Another aspect of this disclosure provides an image processing acceleration system based on a multi-step phase-shifting method, comprising: a driving module for acquiring multiple first image data of a measured object, wherein any two adjacent first image data have a fixed phase difference; and a calculation module for determining the light intensity values of multiple pixels in each of the multiple first image data, and determining an image processing result based on the light intensity values and phase difference of the multiple pixels, wherein the image processing result characterizes the phase information and modulation information of the measured object, and the image processing result is represented by a floating-point number.
[0014] According to an embodiment of this disclosure, the image processing acceleration system further includes: a data caching module connected to the driving module and the computing module, the data caching module consisting of a first-in-first-out memory and a double data rate synchronous dynamic random access memory, used to cache multiple first image data.
[0015] According to an embodiment of this disclosure, the image processing acceleration system further includes: a data transmission module connected to the computing module, used to send the image processing results to a host computer for result verification and data visualization; wherein, the driving module, data caching module, computing module and data transmission module are connected in sequence to form a pipeline structure, and the pipeline structure is used to construct a field-programmable gate array.
[0016] This disclosed method achieves image data acquisition and format conversion, and performs synchronous pixel calculations on multiple image data using the least squares method to obtain accurate image processing results. This enables the method to meet the real-time requirements of high-speed dynamic scenes. This method addresses the significant bottlenecks in processing speed, power consumption, and real-time performance of existing technologies, making it difficult to meet the demands of high-speed acquisition and high frame rate imaging. Attached Figure Description
[0017] The foregoing contents, as well as other objects, features, and advantages of this disclosure, will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:
[0018] Figure 1 A flowchart illustrating an image processing acceleration method based on a multi-step phase-shifting method according to an embodiment of the present disclosure is shown schematically.
[0019] Figure 2 A schematic diagram of a multi-step phase-shifting-based image processing acceleration system according to an embodiment of the present disclosure is shown. Detailed Implementation
[0020] To make the objectives, technical solutions, and advantages of this disclosure clearer, the following detailed description is provided in conjunction with specific embodiments and the accompanying drawings.
[0021] It should be noted that similar or identical parts are referred to by the same reference numerals in the accompanying drawings or description. The technical features of the various embodiments exemplified in the specification can be freely combined to form new solutions without conflict. Furthermore, each claim can stand alone as an embodiment, or the technical features in the various claims can be combined to form new embodiments. In the drawings, the shape or thickness of the embodiments may be enlarged and indicated in a simplified or convenient manner. Moreover, elements or implementations not shown or described in the drawings are those known to those skilled in the art. Additionally, although this document provides examples of parameters containing specific values, it should be understood that the parameters need not be exactly equal to the corresponding values, but can approximate the corresponding values within acceptable error tolerances or design constraints.
[0022] Unless there are technical obstacles or contradictions, the various embodiments described above in this disclosure can be freely combined to form other embodiments, all of which are within the protection scope of this disclosure.
[0023] Although this disclosure has been described in conjunction with the accompanying drawings, the embodiments disclosed in the drawings are intended to illustrate preferred embodiments of this disclosure and should not be construed as limiting the disclosure. The dimensions in the drawings are merely illustrative and should not be construed as limiting the disclosure.
[0024] While some embodiments of the general concept of this disclosure have been shown and described, those skilled in the art will understand that changes may be made to these embodiments without departing from the principles and spirit of the general concept of this disclosure, the scope of which is defined by the claims and their equivalents.
[0025] Figure 1 A flowchart illustrating an image processing acceleration method based on a multi-step phase-shifting method according to an embodiment of the present disclosure is shown. The measurement principle of the multi-step phase-shifting method is to change the phase of the projected light intensity and then acquire the reflected fringe image; that is, to acquire an image after changing the phase, then change it again, and then acquire it again; for example: projecting fringe a1 in step 1, and the camera acquires b1. Projecting fringe a2 in step 2, and the camera acquires b2, until the Nth step image acquisition is completed.
[0026] like Figure 1 As shown, embodiments of this disclosure provide an image processing acceleration method based on a multi-step phase-shifting method, including operations S110~S120.
[0027] In operation S110, multiple first image data of the object under test are acquired, and any two adjacent first image data have a fixed phase difference.
[0028] In some embodiments, multiple first image data can be obtained by sequentially acquiring multiple sinusoidal fringe patterns projected onto the surface of the object under test by a projection device at a fixed phase difference. Then, the multiple sinusoidal fringe patterns are analyzed to obtain multiple second image data, wherein the second image data are represented in fixed-point numbers. Finally, the multiple second image data are decoded to obtain multiple first image data, wherein the first image data are represented in floating-point numbers.
[0029] In some embodiments, fixed-point format conversion represents an operation involving a change in the internal representation of a number. In digital systems, all data is ultimately stored and processed in binary form. Fixed-point format is a method of representing a binary number by pre-defined positions of the decimal point. For example, a 32-bit binary number can be explicitly defined so that the first 16 bits represent the integer part and the last 16 bits represent the fractional part. The advantage of this representation is that it simplifies circuit design, increases processing speed, and consumes less hardware logic resources when performing addition and multiplication operations. Decoding the second image data represented by fixed-point numbers into the first image data represented by floating-point numbers is primarily intended to utilize the first image data represented by floating-point numbers for subsequent calculations, resulting in more accurate image processing results.
[0030] A fixed phase difference means that the phase of the light intensity projected onto the surface of an object by a projection device is constant. This constant phase change relationship is the basis for subsequent quantitative calculations.
[0031] The specific implementation steps of this operation are as follows: First, a structured light projector controlled by a synchronization signal generator can be used to project a series of sinusoidal fringe light fields onto the surface of the object under test at predetermined time intervals. During this process, there is a fixed phase difference between adjacent projected fringe patterns. Then, an industrial camera based on the CoaXPress standard (CXP industrial camera) can be used, under the same synchronization signal control, to capture images of the deformed fringe patterns modulated by the object surface at each stable projection moment to obtain the first image data. For the multi-step phase-shifting method, this process will be repeated multiple times to obtain a sequence of multiple first image data with a fixed phase difference relationship.
[0032] In operation S120, the light intensity values of multiple pixels in each first image data are determined respectively, and the image processing result is determined based on the light intensity values and phase difference of multiple pixels. The image processing result represents the phase information and modulation information of the measured object, and the image processing result is expressed as a floating-point number.
[0033] In some embodiments, determining an image processing result based on the light intensity values and phase differences of multiple pixels includes: determining the vertical offset and horizontal offset corresponding to the phase difference; determining the vertical projection intensity of the light intensity value of each pixel in the vertical direction based on the light intensity value and vertical offset of each pixel in different first image data; determining the horizontal projection intensity of the light intensity value of each pixel in the horizontal direction based on the light intensity value and horizontal offset of each pixel in different first image data; and determining the image processing result based on the vertical projection intensity and horizontal projection intensity of each pixel.
[0034] The formula for calculating ideal light intensity is shown in equation (1):
[0035] (1)
[0036] By performing a mathematical transformation on equation (1), we can obtain equation (2):
[0037] (2)
[0038] in, Let A(x, y) represent the ideal light intensity of a pixel in the nth image data, A(x, y) represent the background light intensity, and B(x, y) represent the ideal modulation intensity of the object being measured. φ represents the phase difference, and φ represents the phase.
[0039] Since the sum of the squares of the differences between the ideal light intensity and the actual light intensity is as shown in equation (3):
[0040] (3)
[0041] Therefore, by using the least squares method to solve the problem, we can obtain:
[0042] (4)
[0043] (5)
[0044] in, This represents the actual light intensity of a pixel in the nth image data. This represents the vertical offset corresponding to the phase difference. This represents the horizontal offset corresponding to the phase difference. The result represents the vertical projection intensity of the light intensity value of each pixel in the vertical direction. The result represents the horizontal projection intensity of the light intensity value of each pixel in the horizontal direction. For the phase information of the object being measured, The modulation information of the object being measured is represented by N, which represents the number of steps. For example, in the eight-step phase-shift method, N=8.
[0045] In some embodiments, the vertical projection intensity of the light intensity value of each pixel in the vertical direction refers to the cumulative result of multiplying the light intensity change of that pixel in a series of images with a fixed phase difference by a series of known vertical offsets. The horizontal projection intensity of the light intensity value of each pixel in the horizontal direction refers to the cumulative result of multiplying the light intensity change of the same pixel by a series of known horizontal offsets.
[0046] The specific calculation steps are as follows: First, read the light intensity value of each pixel in all the first image data. Next, calculate the dot product of the light intensity value with the vertical offset to obtain the vertical projection intensity, and calculate the dot product with the horizontal offset to obtain the horizontal projection intensity. Then, calculate the ratio of the vertical projection intensity to the horizontal projection intensity, and perform an arctangent operation on this ratio. The resulting angle value is the phase information corresponding to that pixel.
[0047] Simultaneously, the sum of the squares of the vertical projection intensity value and the squares of the horizontal projection intensity value is calculated, the square root of this sum is taken, and the result is multiplied by a fixed coefficient determined by the number of phase shift steps to finally obtain the modulation information of the pixel.
[0048] In some embodiments, the entire calculation process of phase information and modulation information is performed using double-precision floating-point format to obtain higher precision phase information and modulation information.
[0049] The multi-step phase-shift-based image processing acceleration method in this embodiment separates and precisely extracts the mixed phase information in the light intensity signal through deterministic projection and vector operation. The calculated modulation information provides a direct verification basis for the reliability of the phase information. The entire algorithm structure is regular and avoids complex iterative calculations, which can effectively improve the overall computing speed and efficiency of the image processing system.
[0050] All the computational operations described above have a corresponding set of computational logic that is executed synchronously in its dedicated pipeline for each pixel position in the image. This can be advanced at a stable clock cycle, with a new batch of pixel data input into the computation pipeline each clock cycle, and a batch of old pixels completing the computation and outputting the results, thus achieving continuous and high-speed processing of the entire image data.
[0051] According to an embodiment of this disclosure, the image processing result is in floating-point format. The image processing acceleration method further includes: converting the image processing result to a fixed-point format to obtain an image processing result represented by a fixed-point number; and sending the image processing result represented by the fixed-point number to a host computer.
[0052] In some embodiments, when image processing results need to be sent to a host computer, the final result is initially also in floating-point format because the acceleration device typically uses floating-point format for computation to improve efficiency. To meet the interface requirements of the host computer software, which usually processes fixed-point data, the phase and modulation result data in floating-point format can be converted to fixed-point numbers. After conversion, data conforming to the standard fixed-point representation specification is obtained. Finally, this fixed-point format data is encapsulated into a specified network or bus protocol data packet and sent to the host computer.
[0053] The image processing acceleration method based on multi-step phase shift in this embodiment of the present disclosure, by performing centralized format conversion, not only ensures the high-speed computing advantage of the hardware, but also ensures the universality and accuracy of the data transmitted to the host computer, avoiding complex format parsing on the software side.
[0054] According to embodiments of this disclosure, the image processing acceleration method further includes: caching a plurality of first image data in a data cache module composed of a first-in-first-out memory and a double data rate synchronous dynamic random access memory; retrieving the plurality of first image data from the data cache module, and determining the light intensity values of a plurality of pixels in each of the first image data.
[0055] In some embodiments, since processing the first image data typically requires a high frame rate (i.e., high-speed processing), all relevant input data can be prepared before starting pixel-level calculations. A data buffer module can be pre-assembled using first-in-first-out (FIFO) memory and double-data-rate synchronous dynamic random access memory. A read operation can then be initiated from this data buffer module. This operation sequentially reads the values of the same pixel coordinates from different pieces of the first image data according to a predetermined address order or addressing mode, forming a sequence of light intensity values for that pixel. This read process ensures that subsequent computation units can obtain all the input data required to process a single pixel at once, thereby enabling continuous pipelined computation.
[0056] The image processing acceleration method based on multi-step phase shifting in this embodiment accurately and promptly delivers image data stored in a high-speed, large-capacity cache to the computing unit through ordered data scheduling, meeting the bandwidth and order requirements of pixel-level parallel processing and avoiding data loss. Based on the image processing acceleration method based on multi-step phase shifting disclosed in the above embodiments, this invention also provides an image processing acceleration system based on multi-step phase shifting, which will be discussed below. Figure 2 The system is described in detail.
[0057] Figure 2 A schematic diagram of an image processing acceleration system based on a multi-step phase-shifting method according to an embodiment of this application is shown.
[0058] like Figure 2 As shown, the image processing acceleration system 200 based on the multi-step phase-shifting method in this embodiment includes a driving module 210, a data cache module 220, a calculation module 230, and a data sending module 240.
[0059] The driving module 210 is used to acquire multiple first image data of the object under test, wherein any two adjacent first image data have a fixed phase difference.
[0060] The data cache module 220 is connected to the driver module 210 and the calculation module 230. The data cache module consists of a first-in-first-out memory and a double data rate synchronous dynamic random access memory, and is used to cache multiple first image data.
[0061] The calculation module 230 is used to determine the light intensity values of multiple pixels in each of the multiple first image data, and to determine the image processing result based on the light intensity values and phase difference of the multiple pixels. The image processing result represents the phase information and modulation information of the measured object, and the image processing result is represented by floating-point numbers.
[0062] The data sending module 240 is connected to the computing module 230 and is used to send the image processing results to the host computer for result verification and data visualization.
[0063] In some embodiments, the driver module 210, data cache module 220, computing module 230, and data transmission module 240 in this embodiment can form a pipelined structure to construct a field-programmable gate array (FPGA). The chip information of the FPGA can be set as follows: Model: XCKU040-2FFVA1156I, Logic Unit: 530250, Flip-flop: 484800, Digital Signal Processor (DSP): 1920, Lookup Table: 242400, Block Random Access Memory (BRAM): 21.1Mb and equipped with four 1GB fourth-generation double data rate synchronous dynamic random access memory (DDR4), with a clock frequency of 200MHz.
[0064] The internal resources of this field-programmable gate array can be utilized in a highly parallel manner, avoiding redundant control overhead, thereby enabling low-latency, high-throughput hardware operations.
[0065] Furthermore, this field-programmable gate array (FPGA) employs an on-demand activation hardware architecture, eliminating the need for instruction decoding and context switching, and automatically controlling hardware power consumption based on data flow. Compared to high-performance graphics processing units (GPUs), this FPGA hardware acceleration module does not require high clock speeds, and the computing units can operate in low-power or sleep states when not in use, significantly reducing system power consumption and making it suitable for portable or mobile measurement devices.
[0066] This implementation uses a pipeline structure formed by the driver module 210, data cache module 220, calculation module 230 and data transmission module 240 to achieve full-process synchronization of image input, caching, calculation and output, which improves the measurement system's ability to respond to dynamic changes, supports continuous processing of each frame of image data, and eliminates the need for redundant intermediate storage.
[0067] It should be noted that the details not covered in the system implementation section are similar to those in the method implementation section, and will not be repeated here.
[0068] It should be understood that the specific order or hierarchy of steps in the disclosed process is an example of an exemplary method. Based on design preferences, it should be understood that the specific order or hierarchy of steps in the process may be rearranged without departing from the scope of this disclosure. The appended method claims provide elements of various steps in an exemplary order and are not intended to limit the scope to a specific order or hierarchy.
[0069] It should also be noted that the directional terms mentioned in the embodiments, such as "up," "down," "front," "back," "left," and "right," are only for reference to the directions in the accompanying drawings and are not intended to limit the scope of protection of this disclosure. Throughout the drawings, the same elements are represented by the same or similar reference numerals. Conventional structures or constructions will be omitted when they may cause confusion in understanding this disclosure. Furthermore, the shapes, sizes, and positional relationships of the components in the drawings do not reflect their actual size, scale, or actual positional relationships.
[0070] In the detailed description above, various features are combined together in a single embodiment to simplify this disclosure. This approach to disclosure should not be construed as reflecting an intention that embodiments of the claimed subject matter require more features than are explicitly stated in each claim. Rather, as reflected in the appended claims, this disclosure is in a state of having fewer features than all of the features of the single disclosed embodiment. Therefore, the appended claims are hereby explicitly incorporated into the detailed description, with each claim representing a separate preferred embodiment of this disclosure.
[0071] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this disclosure, "a plurality of" means at least two, such as two, three, etc., unless otherwise expressly specified. The term "comprising" as used in the specification or claims is interpreted in a manner similar to the term "including," as "including" is used as a conjunction in the claims. The use of any term "or" in the specification or claims is intended to mean "non-exclusive or."
[0072] The specific embodiments described above further illustrate the purpose, technical solutions, and beneficial effects of this disclosure. It should be understood that the above descriptions are merely specific embodiments of this disclosure and are not intended to limit this disclosure. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this disclosure should be included within the protection scope of this disclosure.
Claims
1. An image processing acceleration method based on multi-step phase-shifting, characterized in that, include: Acquire multiple first image data of the object under test, wherein any two adjacent first image data have a fixed phase difference; The light intensity values of multiple pixels in each of the multiple first image data are determined respectively, and the image processing result is determined based on the light intensity values of the multiple pixels and the phase difference. The image processing result represents the phase information and modulation information of the measured object, and the image processing result is expressed as a floating-point number.
2. The image processing acceleration method according to claim 1, characterized in that, The acquisition of multiple first image data of the object under test includes: Multiple sinusoidal stripe patterns are sequentially acquired and projected onto the surface of the object under test by a projection device at a fixed phase difference. The plurality of sinusoidal stripe patterns are analyzed to obtain a plurality of second image data, which are represented by a fixed number of points; The plurality of second image data are decoded to obtain the plurality of first image data, wherein the first image data are represented in floating-point numbers.
3. The image processing acceleration method according to claim 1, characterized in that, Determining the image processing result based on the light intensity values of the plurality of pixels and the phase difference includes: Determine the vertical and horizontal offsets corresponding to the phase difference; Based on the light intensity value of each pixel in different first image data and the vertical offset, determine the vertical projection intensity of the light intensity value of each pixel in the vertical direction. Based on the light intensity value of each pixel in different first image data and the horizontal offset, determine the horizontal projection intensity of the light intensity value of each pixel in the horizontal direction. The image processing result is determined based on the vertical projection intensity and the horizontal projection intensity of each pixel.
4. The image processing acceleration method according to claim 3, characterized in that, Determining the image processing result based on the vertical projection intensity and the horizontal projection intensity of each pixel includes: The phase information of the measured object is obtained by calculating the arctangent function of the ratio of the vertical projection intensity to the horizontal projection intensity. The sum of the squares of the vertical projection intensity and the horizontal projection intensity is calculated by taking the square root and multiplying it by a preset constant scaling factor to obtain the modulation information of the measured object.
5. The image processing acceleration method according to claim 3, characterized in that, The image processing acceleration method further includes: Store the vertical offset and the horizontal offset in a lookup table; Based on the light intensity value of each pixel in different first image data and the lookup table, the vertical projection intensity and horizontal projection intensity of the light intensity value of each pixel are determined.
6. The image processing acceleration method according to claim 1, characterized in that, The image processing acceleration method further includes: The image processing result is converted to a fixed-point format to obtain the image processing result represented in fixed-point numbers; The image processing results, expressed in fixed-point numbers, are sent to the host computer for result verification and visualization.
7. The image processing acceleration method according to claim 1, characterized in that, The method further includes: The plurality of first image data are cached in a data cache module consisting of a first-in-first-out memory and a double data rate synchronous dynamic random access memory; The data cache module retrieves the plurality of first image data and determines the light intensity values of the plurality of pixels in each first image data.
8. An image processing acceleration system based on a multi-step phase-shifting method, characterized in that, include: The driving module is used to acquire multiple first image data of the object under test, wherein any two adjacent first image data have a fixed phase difference. The calculation module is used to determine the light intensity values of multiple pixels in each of the multiple first image data, and to determine the image processing result based on the light intensity values of the multiple pixels and the phase difference. The image processing result represents the phase information and modulation information of the object under test, and the image processing result is expressed as a floating-point number.
9. The image processing acceleration system according to claim 8, characterized in that, The image processing acceleration system also includes: A data caching module, connected to the driving module and the computing module, is composed of a first-in-first-out memory and a double data rate synchronous dynamic random access memory, and is used to cache the plurality of first image data.
10. The image processing acceleration system according to claim 9, characterized in that, The image processing acceleration system also includes: A data sending module, connected to the computing module, is used to send the image processing results to a host computer for result verification and data visualization. The driving module, the data caching module, the computing module, and the data sending module are connected in sequence to form a pipeline structure, which is used to construct a field-programmable gate array.