Image chip verification method and device, equipment, storage medium and program product

Abstract

The invention relates to an image chip verification method and device, computer equipment, a computer readable storage medium and a computer program product. The method comprises the following steps: selecting a part of image slices from a standard image to make a detection operator; sending the detection operator to a C model for processing to obtain golden data of the detection operator; embedding the detection operator into an original image collected by a sensor to obtain a fused image; inputting the fused image into an image chip for processing to obtain a processed image; extracting the detection operator from the processed image to obtain a to-be-detected image; and comparing the image to be detected with the golden data of the detection operator to obtain a verification result of the image chip. Therefore, the test verification of the image chip can be automatically realized under the condition that the image is input in real time, and the verification efficiency of the image chip is remarkably improved.

Description

Technical Field

[0001] The present application relates to the field of image processing technology, and in particular to an image chip verification method, device, computer equipment, computer-readable storage medium and computer program product. Background Art

[0002] With the development of image processing technology, more and more powerful image processor chips have emerged. In the pre-silicon verification process of image processor chips, it is necessary to first verify whether the hardware design of the image processor is correct. Field-Programmable Gate Array (FPGA) prototype verification is a verification technology based on FPGA and is a commonly used verification method in chip design. It verifies the chip by converting the Register Transfer Language (RTL) compiler into a bitfile and transplanting it into the FPGA to verify the correctness of the chip function and timing. In the FPGA verification stage of the chip, it is necessary to write an automated test script to compare the results of the FPGA actual run with the golden sample (golden sample / standard sample) output by the C Model (C language modeling model). All the golden data (golden data) of each module must be verified to meet the acceptance criteria.

[0003] Currently, the dynamic input image data can only be detected by naked eye. However, naked eye detection cannot realize automated test verification, especially when the frame rate is fast, it is impossible to accurately judge whether each frame is abnormal, and it is even more impossible to judge pixel abnormalities. In the chip verification stage, there is no good way to realize automated testing under dynamic image input. Summary of the invention

[0004] Based on this, it is necessary to provide an image chip verification method, device, computer equipment, computer-readable storage medium and computer program product that can realize automatic chip testing under real-time image input to address the above technical problems.

[0005] In a first aspect, the present application provides an image chip verification method, comprising:

[0006] Select a portion of image slices from the standard image to create a detection operator;

[0007] Send the detection operator to the C model for processing to obtain the golden data of the detection operator;

[0008] Embedding the detection operator into the original image collected by the sensor to obtain a fused image;

[0009] Inputting the fused image into an image chip for processing to obtain a processed image;

[0010] Extracting the detection operator from the processed image to obtain an image to be detected;

[0011] The image to be detected is compared with the golden data of the detection operator to obtain the verification result of the image chip.

[0012] In one embodiment, a portion of image slices is selected from a standard image to produce a detection operator, including:

[0013] Divide the standard image into X image slices of the same size, wherein the size of the standard image is consistent with the size of the original image collected by the sensor; X is a natural number greater than 1;

[0014] At least one image slice is selected from X image slices of the same size as a detection operator.

[0015] In one embodiment, the detection operator is embedded into the original image collected by the sensor to obtain a fused image, including:

[0016] An implicit fusion method is used to transform the detection operator into different domains, and an implicit embedding method is used to embed the detection operator into the original image collected by the sensor to obtain a fused image.

[0017] In one embodiment, the detection operator is transformed into a different domain by an implicit fusion method, and the detection operator is embedded into the original image collected by the sensor by an implicit embedding method to obtain a fused image, including:

[0018] The original image collected by the sensor is used as host information for preprocessing and then subjected to discrete wavelet transformation to obtain a first low-frequency component and a first high-frequency component;

[0019] Select the first low-frequency component for singular value decomposition to obtain the host's left singular matrix U, the host's singular value matrix Σ, and the host's right singular matrix V;

[0020] Performing singular value decomposition on the detection operator to obtain a singular value matrix of the detection operator;

[0021] Add the singular value matrix of the detection operator to the singular value matrix Σ of the host to obtain a new singular value matrix Σ';

[0022] The new singular value matrix Σ' is combined with the left singular matrix U of the host and the right singular matrix V of the host to perform an inverse singular value decomposition transform to obtain a low-frequency component embedded with a detection operator;

[0023] The low-frequency component embedded with the detection operator and the first high-frequency component are subjected to inverse discrete wavelet transform to obtain a fused image.

[0024] In one embodiment, extracting the detection operator from the processed image to obtain the image to be detected includes:

[0025] Performing discrete wavelet transform on the fused image to obtain a second low-frequency component and a second high-frequency component;

[0026] Selecting the second low frequency component to perform singular value decomposition, and estimating the singular value matrix of the processed image according to the decomposition result and the singular value matrix of the processed image;

[0027] According to the estimated singular value matrix of the processed image, a left singular matrix of the processed image and a right singular matrix of the processed image are combined to perform an inverse singular value decomposition transform to obtain extracted detection operator data;

[0028] The extracted detection operator data is verified, and if the verification passes, the image to be detected is obtained.

[0029] In one embodiment, when the original image captured by the sensor is a continuous N-frame image, N is a natural number greater than 1, the method further includes:

[0030] Select image slices at different positions in the standard image to respectively create corresponding detection operators, and obtain M detection operators; where M is a natural number greater than 1 and less than or equal to N;

[0031] Detection operators are randomly selected from the M detection operators, and are respectively embedded in different positions of N frames of images to obtain N frames of fused images.

[0032] In one embodiment, the image to be detected is compared with the golden data of the detection operator to obtain the verification result of the image chip, including:

[0033] Using a quantitative evaluation method to compare the image to be detected with the golden data of the detection operator to obtain an image comparison result, wherein the comparison result includes: any one of similarity, difference, and quality index;

[0034] The quantitative evaluation method includes: any one of normalization coefficient, histogram comparison, feature point matching, and structural similarity index.

[0035] In a second aspect, the present application also provides an image chip verification device, comprising:

[0036] A selection module is used to select a portion of image slices from a standard image to make a detection operator;

[0037] A golden data acquisition module, used for sending the detection operator to the C model for processing to obtain the golden data of the detection operator;

[0038] A fusion module, used for embedding the detection operator into the original image collected by the sensor to obtain a fused image;

[0039] A processing module, used for inputting the fused image into an image chip for processing to obtain a processed image;

[0040] An extraction module, used to extract the detection operator from the processed image to obtain an image to be detected;

[0041] The verification module is used to compare the image to be detected with the golden data of the detection operator to obtain the verification result of the image chip.

[0042] In a third aspect, the present application further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0043] Select a portion of image slices from the standard image to create a detection operator;

[0044] Sending the detection operator to the C model for processing to obtain golden data of the detection operator;

[0045] Embedding the detection operator into the original image collected by the sensor to obtain a fused image;

[0046] Inputting the fused image into an image chip for processing to obtain a processed image;

[0047] Extracting the detection operator from the processed image to obtain an image to be detected;

[0048] The image to be detected is compared with the golden data of the detection operator to obtain the verification result of the image chip.

[0049] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0050] Select a portion of image slices from the standard image to create a detection operator;

[0051] Sending the detection operator to the C model for processing to obtain golden data of the detection operator;

[0052] Embedding the detection operator into the original image collected by the sensor to obtain a fused image;

[0053] Inputting the fused image into an image chip for processing to obtain a processed image;

[0054] Extracting the detection operator from the processed image to obtain an image to be detected;

[0055] The image to be detected is compared with the golden data of the detection operator to obtain the verification result of the image chip.

[0056] In a fifth aspect, the present application further provides a computer program product, including a computer program, which implements the following steps when executed by a processor:

[0057] Select a portion of image slices from the standard image to create a detection operator;

[0058] Sending the detection operator to the C model for processing to obtain golden data of the detection operator;

[0059] Embedding the detection operator into the original image collected by the sensor to obtain a fused image;

[0060] Inputting the fused image into an image chip for processing to obtain a processed image;

[0061] Extracting the detection operator from the processed image to obtain an image to be detected;

[0062] The image to be detected is compared with the golden data of the detection operator to obtain the verification result of the image chip.

[0063] The above-mentioned image chip verification method, device, computer equipment, computer-readable storage medium and computer program product make a detection operator by selecting a part of image slices from a standard image; thus, it can be applied to the situation where there is no fixed scene and the standard reference picture cannot be collected when the image is input in real time. The detection operator is sent to the C model for processing to obtain the golden data of the detection operator; the detection operator is embedded in the original image collected by the sensor to obtain the fused image; thus, the image fusion can be quickly realized. The fused image is input to the image chip for processing to obtain the processed image; the detection operator is extracted from the processed image to obtain the image to be detected; thus, the image containing the detection operator can be input to the image chip for processing to verify whether the design of the image chip is correct through the detection operator. The image to be detected is compared with the golden data of the detection operator to obtain the verification result of the image chip. Thus, the test verification of the image chip can be automatically realized when the image is input in real time, which significantly improves the verification efficiency of the image chip. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0065] FIG1 is a schematic diagram of the overall process of image chip design and verification in one embodiment;

[0066] FIG2 is a schematic diagram of a flow chart of an image chip verification method in one embodiment;

[0067] FIG3 is a schematic diagram of the manufacturing principle of a detection operator in one embodiment;

[0068] FIG4 is a schematic diagram of the principle of image fusion in one embodiment;

[0069] FIG5 is a schematic diagram of the principle of image extraction in one embodiment;

[0070] FIG6 is a schematic diagram of the effect of detection operator extraction in one embodiment;

[0071] FIG7 is a schematic diagram of a flow chart of an image chip verification method in another embodiment;

[0072] FIG8 is a schematic diagram of a fusion process of traversing original images in a sliding window manner in one embodiment;

[0073] FIG9 is a structural block diagram of an image chip verification device in one embodiment;

[0074] FIG. 10 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0075] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0076] The image chip verification method provided in the embodiment of the present application can be used to realize the automated test verification of the image signal processing (Image Single Processor, ISP) chip under the condition of real-time input image, so as to improve the verification efficiency and coverage. FIG1 shows the overall process of the pre-silicon verification process of the image processor chip, which can specifically include: demand analysis, demand definition, architecture design, chip modeling, algorithm modeling, C Model (model), logic design (front-end design, register transfer level RTL design), netlist file, FPGA (consistency verification), physical design (back-end design), TO (Tape Out, tape-out), and return sample. Among them, when the FPGA and the C Model read the same input raw image and the same isp parameters, the same results will be generated in theory. Therefore, as long as the output of the FPGA is compared with the golden data generated by the C Model, the correctness of the image chip function and timing can be verified.

[0077] In an exemplary embodiment, as shown in FIG2 , an image chip verification method is provided, and the method is applied to an image chip test scenario, including the following steps 201 to 205. Among them:

[0078] Step 201 , select a portion of image slices from a standard image to create a detection operator.

[0079] The standard image may be a raw image of a standard color card captured by a sensor. A raw format image refers to the original data obtained by converting a captured light source signal into a digital signal by an image sensor.

[0080] Step 202: Send the detection operator to the C model for processing to obtain the golden data of the detection operator.

[0081] Exemplarily, FIG3 shows the process of generating golden data. As shown in FIG3, a standard color card is used as input (the raw image corresponding to the standard color card is collected by the sensor), and the size of the input source raw image is consistent with the size of the original image collected by the sensor. The golden data is marked into X image slices of the same size according to the size of m*n, and the coordinates and size of the area are recorded. The raw image corresponding to the standard color card is input into the C model to obtain the golden data, and the obtained golden data is marked to obtain several m*n sized areas corresponding to the image slices, so that the golden data corresponding to each area can be obtained.

[0082] In this embodiment, a raw image corresponding to a frame of a standard color card collected by a sensor can be input into a C language modeling model (cmodel, also referred to as C model) to obtain golden data, and the golden data can be divided into multiple slices of the same size, which correspond one-to-one to the multiple slices divided by the raw image, as the golden of each image slice in the raw image.

[0083] Exemplarily, a standard image is divided into X image slices of the same size, wherein the size of the standard image is consistent with the size of the original image acquired by the sensor; X is a natural number greater than 1; and at least one image slice is selected from the X image slices of the same size as a detection operator.

[0084] Step 203: embed the detection operator into the original image collected by the sensor to obtain a fused image.

[0085] Exemplarily, an implicit fusion method may be used to transform the detection operator into a different domain, and an implicit embedding method may be used to embed the watermark information contained in the detection operator into the original image collected by the sensor to obtain a fused image.

[0086] FIG4 shows the image fusion process. As shown in FIG4, various implicit extraction algorithms can be used to transform the detection operator to other domains. For example, the domain change of the detection operator can be performed based on discrete cosine transform (DCT), Riemann transform (DFT), discrete wavelet transform (DWT), Harris-SIFT feature region, and multi-domain transform based on strong robustness. Then, the transformed detection operator is embedded into the original image collected by the sensor using an implicit embedding method.

[0087] Exemplarily, taking discrete wavelet transform as an example, the original image collected by the sensor is first preprocessed as host information and then subjected to discrete wavelet transform to obtain a first low-frequency component and a first high-frequency component; the first low-frequency component is selected for singular value decomposition to obtain the host's left singular matrix U, the host's singular value matrix Σ, and the host's right singular matrix V; the detection operator is subjected to singular value decomposition to obtain the singular value matrix of the detection operator; the singular value matrix of the detection operator is added to the host's singular value matrix Σ to obtain a new singular value matrix Σ'; the new singular value matrix Σ' is combined with the host's left singular matrix U and the host's right singular matrix V to perform singular value decomposition inverse transform to obtain a low-frequency component embedded with the detection operator; the low-frequency component embedded with the detection operator and the first high-frequency component are subjected to discrete wavelet inverse transform to obtain a fused image.

[0088] In this embodiment, the original image is used as the host signal and subjected to DWT (discrete wavelet) transformation to decompose it into different frequency components, usually including a low-frequency component (LL) and three high-frequency components (LH, HL, HH); the low-frequency component is selected for SVD (singular value decomposition) to obtain a left singular matrix (U), a singular value matrix (Σ), and a right singular matrix (V). The detection operator is subjected to SVD processing to obtain its corresponding three matrices. The detection operator is embedded by adjusting the singular values ​​of the low-frequency components of the host signal (usually involving adding the singular value matrix Σ of the detection operator to the singular value matrix Σ of the host signal to form a new singular value matrix Σ'). Using the new singular value matrix Σ', combined with the U matrix and V matrix of the original image, an SVD inverse transformation is performed to obtain the low-frequency component embedded with the detection operator. The modified low-frequency component is subjected to an inverse DWT transformation together with the original three high-frequency components to obtain the final host signal containing the detection operator (i.e., the fused image).

[0089] Step 204: input the fused image into an image chip for processing to obtain a processed image.

[0090] In this embodiment, the fused image is input to ISP (image signal processing) for processing, and the processed image is output, wherein the coordinates of the standard operator in the processed image are recorded for subsequent extraction processing.

[0091] Step 205: extract the detection operator from the processed image to obtain the image to be detected.

[0092] The image extraction process is shown in FIG5 . As shown in FIG5 , the extraction method is determined according to the fusion method selection. The inverse transformation corresponding to the fusion method is used to extract the detection operator.

[0093] Exemplarily, taking discrete wavelet transform DWT as an example, firstly, the fused image is subjected to discrete wavelet transform to obtain a second low-frequency component and a second high-frequency component; the second low-frequency component is selected for singular value decomposition, and the singular value matrix of the processed image is estimated based on the decomposition result and the singular value matrix of the processed image; based on the estimated singular value matrix of the processed image, the left singular matrix of the processed image and the right singular matrix of the processed image are combined to perform inverse singular value decomposition transform to obtain extracted detection operator data; the extracted detection operator data is verified, and if the verification passes, the image to be detected is obtained.

[0094] In this embodiment, the original image is used as the host signal, and the host signal containing the detection operator is transformed again by DWT to obtain the low-frequency component. The obtained low-frequency component is decomposed by SVD to obtain U matrix, Σ1 matrix, and V matrix. By comparing Σ1 with the singular value matrix Σ of the original host signal, the difference is calculated, and then the singular value matrix Σw of the processed image is estimated. Using the estimated Σw, combined with the Uw and Vw matrices of the watermark, an SVD inverse transform is performed to restore the watermark information. The extracted watermark information is verified to ensure its correctness and completeness. As an example, Figure 6 shows the watermark extraction result of the watermarked image using the method in this embodiment.

[0095] Step 206 , compare the image to be detected with the golden data of the detection operator to obtain the verification result of the image chip.

[0096] In this embodiment, the image to be detected can be compared with the golden data of the detection operator by using a quantitative evaluation method to obtain the image comparison result, wherein the comparison result includes: similarity, difference, and any one of the quality indicators; the quantitative evaluation method includes: normalization coefficient, histogram comparison, feature point matching, and any one of the structural similarity index. Thus, a more accurate verification result can be obtained.

[0097] The method in this embodiment selects a portion of image slices from a standard image to make a detection operator; thus, it can be applied to situations where there is no fixed scene and the standard reference image cannot be collected when the image is input in real time. The detection operator is sent to the C model for processing to obtain the golden data of the detection operator; the detection operator is embedded in the original image collected by the sensor to obtain a fused image; thus, the image fusion can be quickly realized. The fused image is input to the image chip for processing to obtain a processed image; the detection operator is extracted from the processed image to obtain the image to be detected; thus, the image containing the detection operator can be input to the image chip for processing to verify whether the design of the image chip is correct through the detection operator. The image to be detected is compared with the golden data of the detection operator to obtain the verification result of the image chip. Thus, the test verification of the image chip can be automatically realized when the image is input in real time, which significantly improves the verification efficiency of the image chip. In addition, since each detection only needs to detect a smaller image area, the detection efficiency can be significantly improved, and the robustness is good.

[0098] In another exemplary embodiment, as shown in FIG7 , an image chip verification method is provided, and the method is applied to an image chip test scenario, including the following steps 701 to 705. Among them:

[0099] Step 701: When the original image collected by the sensor is a continuous N-frame image, image slices at different positions in the standard image are selected to respectively create corresponding detection operators to obtain M detection operators.

[0100] In this embodiment, N is a natural number greater than 1, and M is a natural number greater than 1 and less than or equal to N.

[0101] Step 702: Send the M detection operators to the C model for processing to obtain golden data of the M detection operators.

[0102] Step 703 , randomly select detection operators from the M detection operators, and embed them into different positions of the N frames of images respectively, to obtain N frames of fused images.

[0103] In this embodiment, since the sensor will continuously collect N frames of images in the real-time input scenario, by fusing the detection operators in different areas of these N frames (the selected detection operators are also image slices at different positions), recording the coordinates of the fused detection operators and the fused position information in the collected images, the corresponding fused area can be extracted during extraction, so that all input areas of the sensor can be traversed through N frames of images.

[0104] Exemplarily, as shown in FIG8 , a sliding window approach may be used to allow different detection operators to traverse all positions of the image, so that the entire image can be detected.

[0105] Step 704: input the fused image into an image chip for processing to obtain a processed image.

[0106] Step 705: extract the detection operator from the processed image to obtain the image to be detected.

[0107] Step 706, compare the image to be detected with the golden data of the detection operator to obtain the verification result of the image chip.

[0108] For the specific implementation process and technical effects of steps 704 to 706 in this embodiment, please refer to the relevant descriptions of steps 204 to 206 in the method shown in FIG. 2 , which will not be repeated here.

[0109] In this embodiment, when the original image collected by the sensor is a continuous N-frame image, image slices at different positions in the standard image are selected to respectively create corresponding detection operators, thereby obtaining M detection operators. Detection operators are randomly selected from the M detection operators and embedded in different positions of the N-frame images to obtain N-frame fused images. In this way, different areas of the image can be tested and verified, making the verification results more comprehensive and accurate.

[0110] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0111] Based on the same inventive concept, the embodiment of the present application also provides an image chip verification device for implementing the image chip verification method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more image chip verification device embodiments provided below can refer to the limitations of the image chip verification method above, and will not be repeated here.

[0112] In an exemplary embodiment, as shown in FIG9 , an image chip verification device is provided, comprising: a selection module 901, a golden data acquisition module 902, a fusion module 903, a processing module 904, an extraction module 905 and a verification module 906, wherein:

[0113] A selection module 901 is used to select a portion of image slices from a standard image to produce a detection operator;

[0114] Golden data acquisition module 902, used for sending the detection operator to C model for processing to obtain golden data of the detection operator;

[0115] A fusion module 903 is used to embed the detection operator into the original image collected by the sensor to obtain a fused image;

[0116] A processing module 904 is used to input the fused image into an image chip for processing to obtain a processed image;

[0117] An extraction module 905 is used to extract the detection operator from the processed image to obtain an image to be detected;

[0118] The verification module 906 is used to compare the image to be detected with the golden data of the detection operator to obtain the verification result of the image chip.

[0119] Exemplarily, the selection module 901 is specifically used to: divide the standard image into X image slices of the same size, wherein the size of the standard image is consistent with the size of the original image collected by the sensor; X is a natural number greater than 1; and select at least one image slice from the X image slices of the same size as a detection operator.

[0120] Exemplarily, the fusion module 903 is specifically used to transform the detection operator into a different domain by using an implicit fusion method, and embed the detection operator into the original image collected by the sensor by using an implicit embedding method to obtain a fused image.

[0121] Exemplarily, the fusion module 903 is specifically used to preprocess the original image collected by the sensor as the host information and then perform discrete wavelet transform to obtain a first low-frequency component and a first high-frequency component; select the first low-frequency component for singular value decomposition to obtain the host's left singular matrix U, the host's singular value matrix Σ, and the host's right singular matrix V; perform singular value decomposition on the detection operator to obtain the singular value matrix of the detection operator; add the singular value matrix of the detection operator to the host's singular value matrix Σ to obtain a new singular value matrix Σ'; perform singular value decomposition inverse transform on the new singular value matrix Σ' in combination with the host's left singular matrix U and the host's right singular matrix V to obtain a low-frequency component embedded with the detection operator; perform discrete wavelet inverse transform on the low-frequency component embedded with the detection operator and the first high-frequency component to obtain a fused image.

[0122] Exemplarily, the extraction module 905 is specifically used to perform discrete wavelet transform on the fused image to obtain a second low-frequency component and a second high-frequency component; select the second low-frequency component for singular value decomposition, and estimate the singular value matrix of the processed image based on the decomposition result and the singular value matrix of the processed image; perform inverse singular value decomposition transform based on the estimated singular value matrix of the processed image in combination with the left singular matrix of the processed image and the right singular matrix of the processed image to obtain extracted detection operator data; verify the extracted detection operator data, and if the verification passes, obtain the image to be detected.

[0123] Exemplarily, when the original image collected by the sensor is a continuous N-frame image, the selection module 901 is further used to select image slices at different positions in the standard image to respectively create corresponding detection operators, and obtain M detection operators; wherein M is a natural number greater than 1 and less than or equal to N. The fusion module 903 is further used to randomly select detection operators from the M detection operators, and embed them into different positions of the N-frame image, respectively, to obtain an N-frame fused image.

[0124] Exemplarily, the verification module 906 is specifically used to compare the image to be detected with the golden data of the detection operator in a quantitative evaluation manner to obtain an image comparison result, wherein the comparison result includes: any one of similarity, difference, and quality index;

[0125] The quantitative evaluation method includes: any one of normalization coefficient, histogram comparison, feature point matching, and structural similarity index.

[0126] In this embodiment, a detection operator is made by selecting a part of image slices from a standard image; thus, it can be applied to situations where there is no fixed scene and a standard reference image cannot be collected when the image is input in real time. The detection operator is embedded in the original image collected by the sensor to obtain a fused image; thus, image fusion can be quickly realized. The fused image is input into the image chip for processing to obtain a processed image; the detection operator is extracted from the processed image to obtain an image to be detected; thus, the image containing the detection operator can be input into the image chip for processing to verify whether the design of the image chip is correct through the detection operator. The image to be detected is compared with the golden data of the detection operator to obtain the verification result of the image chip. Thus, the test and verification of the image chip can be automatically realized when the image is input in real time, which significantly improves the verification efficiency of the image chip.

[0127] Each module in the above-mentioned image chip verification device can be implemented in whole or in part by software, hardware or a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0128] In an exemplary embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be shown in FIG10. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit, and an input device. The processor, the memory, and the input / output interface are connected via a system bus, and the communication interface, the display unit, and the input device are connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner may be implemented through WIFI, a mobile cellular network, near field communication (NFC) or other technologies. When the computer program is executed by the processor, an image chip verification method is implemented. The display unit of the computer device is used to form a visually visible picture, which may be a display screen, a projection device, or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.

[0129] Those skilled in the art will understand that the structure shown in FIG. 10 is merely a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0130] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

[0131] Select a portion of image slices from the standard image to create a detection operator;

[0132] Sending the detection operator to the C model for processing to obtain golden data of the detection operator;

[0133] Embedding the detection operator into the original image collected by the sensor to obtain a fused image;

[0134] Inputting the fused image into an image chip for processing to obtain a processed image;

[0135] Extracting the detection operator from the processed image to obtain an image to be detected;

[0136] The image to be detected is compared with the golden data of the detection operator to obtain the verification result of the image chip.

[0137] In one embodiment, when the processor executes the computer program, the following steps are also implemented:

[0138] A part of image slices is selected from a standard image to produce a detection operator, including: dividing the standard image into X image slices of the same size, wherein the size of the standard image is consistent with the size of an original image collected by a sensor; X is a natural number greater than 1; and selecting at least one image slice from the X image slices of the same size as a detection operator.

[0139] In one embodiment, when the processor executes the computer program, the following steps are also implemented:

[0140] The detection operator is embedded into the original image collected by the sensor to obtain a fused image, including: using an implicit fusion method to transform the detection operator into a different domain, and using an implicit embedding method to embed the detection operator into the original image collected by the sensor to obtain a fused image.

[0141] In one embodiment, when the processor executes the computer program, the following steps are also implemented:

[0142] The detection operator is transformed into a different domain by an implicit fusion method, and the detection operator is embedded into the original image collected by the sensor by an implicit embedding method to obtain a fused image, including:

[0143] The original image collected by the sensor is used as host information for preprocessing and then discrete wavelet transform is performed to obtain a first low-frequency component and a first high-frequency component; the first low-frequency component is selected for singular value decomposition to obtain the host's left singular matrix U, the host's singular value matrix Σ, and the host's right singular matrix V; the detection operator is subjected to singular value decomposition to obtain the detection operator's singular value matrix; the detection operator's singular value matrix is ​​added to the host's singular value matrix Σ to obtain a new singular value matrix Σ'; the new singular value matrix Σ' is combined with the host's left singular matrix U and the host's right singular matrix V to perform singular value decomposition inverse transform to obtain a low-frequency component embedded with the detection operator; the low-frequency component embedded with the detection operator and the first high-frequency component are subjected to discrete wavelet inverse transform to obtain a fused image.

[0144] In one embodiment, when the processor executes the computer program, the following steps are also implemented:

[0145] Extracting the detection operator from the processed image to obtain an image to be detected includes:

[0146] The fused image is subjected to discrete wavelet transform to obtain a second low-frequency component and a second high-frequency component; the second low-frequency component is selected for singular value decomposition, and the singular value matrix of the processed image is estimated according to the decomposition result and the singular value matrix of the processed image; based on the estimated singular value matrix of the processed image, a left singular matrix of the processed image and a right singular matrix of the processed image are combined to perform an inverse singular value decomposition transform to obtain extracted detection operator data; the extracted detection operator data is verified, and if the verification is passed, the image to be detected is obtained.

[0147] In one embodiment, when the processor executes the computer program, the following steps are also implemented:

[0148] When the original image collected by the sensor is a continuous N-frame image, where N is a natural number greater than 1, the method further includes:

[0149] Image slices at different positions in the standard image are selected to respectively create corresponding detection operators to obtain M detection operators; wherein M is a natural number greater than 1 and less than or equal to N; detection operators are randomly selected from the M detection operators, and are respectively embedded in different positions of N frames of images to obtain N frames of fused images.

[0150] In one embodiment, when the processor executes the computer program, the following steps are also implemented:

[0151] Compare the image to be detected with the golden data of the detection operator to obtain the verification result of the image chip, including:

[0152] Using a quantitative evaluation method to compare the image to be detected with the golden data of the detection operator to obtain an image comparison result, wherein the comparison result includes: any one of similarity, difference, and quality index;

[0153] The quantitative evaluation method includes: any one of normalization coefficient, histogram comparison, feature point matching, and structural similarity index.

[0154] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0155] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above-mentioned method embodiments when executed by a processor.

[0156] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0157] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0158] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0159] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. An image chip verification method, characterized in that: The method comprises: Select a portion of image slices from the standard image to create a detection operator; Sending the detection operator to the C model for processing to obtain golden data of the detection operator; Embedding the detection operator into the original image collected by the sensor to obtain a fused image; Inputting the fused image into an image chip for processing to obtain a processed image; Extracting the detection operator from the processed image to obtain an image to be detected; The image to be detected is compared with the golden data of the detection operator to obtain the verification result of the image chip.

2. The method according to claim 1, characterized in that The step of selecting a portion of image slices from a standard image to create a detection operator includes: Divide the standard image into X image slices of the same size, wherein the size of the standard image is consistent with the size of the original image collected by the sensor; X is a natural number greater than 1; At least one image slice is selected from X image slices of the same size as a detection operator.

3. The method according to claim 1, characterized in that The step of embedding the detection operator into the original image collected by the sensor to obtain a fused image includes: An implicit fusion method is used to transform the detection operator into different domains, and an implicit embedding method is used to embed the detection operator into the original image collected by the sensor to obtain a fused image.

4. The method according to claim 3, characterized in that The detection operator is transformed into a different domain by an implicit fusion method, and the detection operator is embedded into the original image collected by the sensor by an implicit embedding method to obtain a fused image, including: The original image collected by the sensor is used as host information for preprocessing and then subjected to discrete wavelet transformation to obtain a first low-frequency component and a first high-frequency component; Select the first low-frequency component for singular value decomposition to obtain the host's left singular matrix U, the host's singular value matrix Σ, and the host's right singular matrix V; Performing singular value decomposition on the detection operator to obtain a singular value matrix of the detection operator; Add the singular value matrix of the detection operator to the singular value matrix Σ of the host to obtain a new singular value matrix Σ'; The new singular value matrix Σ' is combined with the left singular matrix U of the host and the right singular matrix V of the host to perform an inverse singular value decomposition transform to obtain a low-frequency component embedded with a detection operator; The low-frequency component embedded with the detection operator and the first high-frequency component are subjected to inverse discrete wavelet transform to obtain a fused image.

5. The method according to claim 4, characterized in that Extracting the detection operator from the processed image to obtain an image to be detected includes: Performing discrete wavelet transform on the fused image to obtain a second low-frequency component and a second high-frequency component; Selecting the second low frequency component to perform singular value decomposition, and estimating the singular value matrix of the processed image according to the decomposition result and the singular value matrix of the processed image; According to the estimated singular value matrix of the processed image, a left singular matrix of the processed image and a right singular matrix of the processed image are combined to perform an inverse singular value decomposition transform to obtain extracted detection operator data; The extracted detection operator data is verified, and if the verification passes, the image to be detected is obtained.

6. The method according to any one of claims 1 to 5, characterized in that: When the original image collected by the sensor is a continuous N-frame image, where N is a natural number greater than 1, the method further includes: Select image slices at different positions in the standard image to respectively create corresponding detection operators, and obtain M detection operators; where M is a natural number greater than 1 and less than or equal to N; Detection operators are randomly selected from the M detection operators, and are respectively embedded in different positions of N frames of images to obtain N frames of fused images.

7. The method according to any one of claims 1 to 5, characterized in that: Compare the image to be detected with the golden data of the detection operator to obtain the verification result of the image chip, including: Using a quantitative evaluation method to compare the image to be detected with the golden data of the detection operator to obtain an image comparison result, wherein the comparison result includes: any one of similarity, difference, and quality index; The quantitative evaluation method includes: any one of normalization coefficient, histogram comparison, feature point matching, and structural similarity index.

8. An image chip verification device, characterized in that: The device comprises: A selection module is used to select a portion of image slices from a standard image to make a detection operator; A golden data acquisition module, used for sending the detection operator to the C model for processing to obtain the golden data of the detection operator; A fusion module, used for embedding the detection operator into the original image collected by the sensor to obtain a fused image; A processing module, used for inputting the fused image into an image chip for processing to obtain a processed image; An extraction module, used to extract the detection operator from the processed image to obtain an image to be detected; The verification module is used to compare the image to be detected with the golden data of the detection operator to obtain the verification result of the image chip.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

11. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images