A method and system for analyzing impurities in basic nickel carbonate

Abstract

This invention discloses a method and system for analyzing impurities in basic nickel carbonate. The method includes: determining whether each first identification result is an abnormal identification result; if a first identification result is an abnormal identification result, acquiring a target original image corresponding to the first identification result, determining a position of the target original image in a sequence of original images, selecting other original images from other sequences based on the position; inputting the other original images into an image recognition model, the image recognition model outputting a second identification result corresponding to the other original images, and using a preset impurity analysis strategy to perform basic nickel carbonate impurity analysis on each original image in the first and other sequences, obtaining the analysis results of basic nickel carbonate on each production line. This method can effectively improve the efficiency of basic nickel carbonate impurity analysis throughout the entire production line.

Description

Technical Field

[0001] This invention belongs to the field of basic nickel carbonate treatment technology, and particularly relates to a method and system for analyzing impurities in basic nickel carbonate. Background Technology

[0002] In the production of basic nickel carbonate, the purity of the nickel carbonate directly affects the quality of the final product. However, due to the complexity of the production process, equipment wear and tear, or process fluctuations, impurities such as metal particles and reactor debris may be introduced into the basic nickel carbonate. These impurities not only reduce product quality but may also cause subsequent process failures or even lead to the scrapping of the entire batch, resulting in significant economic losses.

[0003] Currently, the following methods are mainly used in industry for impurity detection:

[0004] Manual visual inspection: This involves workers taking samples periodically for observation, but this method is inefficient and greatly affected by subjective factors, making it difficult to detect minute impurities.

[0005] Traditional image processing relies on pixel analysis based on fixed thresholds (such as edge detection and grayscale contrast), but it is easily affected by changes in lighting and device reflections, resulting in a high false alarm rate.

[0006] Static machine learning models: These models use pre-trained CNNs for single-frame image classification, but they cannot correlate with temporal changes and are difficult to distinguish between transient noise and continuous contamination.

[0007] The existing technology has the following drawbacks:

[0008] Insufficient sensitivity: Traditional algorithms are slow to respond to small impurities in the early stages and are often only detected after the contamination has spread.

[0009] Cross-production line traceability is difficult: When multiple production lines are operating in parallel, it is difficult to quickly locate the source of pollution caused by shared equipment (such as conveying pipelines and branch valves).

[0010] Therefore, there is an urgent need for a basic nickel carbonate impurity analysis method that can detect early impurities with high sensitivity and accurately locate pollution sources across production lines, in order to meet the quality control requirements of modern industrial production. Summary of the Invention

[0011] This invention provides a method and system for analyzing impurities in basic nickel carbonate, which solves the technical problem of difficulty in quickly locating pollution sources caused by shared equipment when multiple production lines are operating in parallel.

[0012] In a first aspect, the present invention provides a method for analyzing impurities in basic nickel carbonate, comprising:

[0013] The original images of basic nickel carbonate on each production line are acquired within a preset time period, and the original images are sorted by time based on different production lines to obtain at least one original image sequence.

[0014] Obtain at least one image change rate of the original images in the same production line within a preset time period. Based on the at least one image change rate, use a preset image extraction strategy to extract images from each original image sequence to obtain a target original image set corresponding to each original image sequence.

[0015] Each original target image in a set of original target images is input into a preset image recognition model, and the image recognition model outputs a first recognition result corresponding to each original target image;

[0016] Determine whether each first identification result is an abnormal identification result, wherein the abnormal identification result is the identification result of impurities in basic nickel carbonate;

[0017] If a certain first identification result is an abnormal identification result, then a certain target original image corresponding to the certain first identification result is obtained, and a certain position of the certain target original image in a certain original image sequence is determined. Other original images are selected from other original image sequences based on the certain position, wherein the other original image sequences are the original image sequences after removing a certain original image sequence from all original image sequences.

[0018] The other original images are input into the image recognition model, and the image recognition model outputs a second recognition result corresponding to the other original images. Based on the second recognition result and the first recognition result, a preset impurity analysis strategy is used to perform basic nickel carbonate impurity analysis on the original image sequence and each original image in the other original image sequences to obtain the analysis results of basic nickel carbonate on each production line.

[0019] In a second aspect, the present invention provides a basic nickel carbonate impurity analysis system, comprising:

[0020] The sorting module is configured to acquire the original images of basic nickel carbonate on each production line within a preset time period, and sort the original images by time based on different production lines to obtain at least one original image sequence.

[0021] The extraction module is configured to obtain at least one image change rate of the original images in the same production line within a preset time period, and extract images from each original image sequence according to the at least one image change rate using a preset image extraction strategy to obtain a target original image set corresponding to each original image sequence.

[0022] The first output module is configured to input each target original image from a set of target original images into a preset image recognition model, and the image recognition model outputs a first recognition result corresponding to each target original image;

[0023] The judgment module is configured to judge whether each first identification result is an abnormal identification result, wherein the abnormal identification result is the identification result of impurities in basic nickel carbonate;

[0024] The determination module is configured to, if a certain first recognition result is an abnormal recognition result, acquire a certain target original image corresponding to the certain first recognition result, determine a certain position of the certain target original image in a certain original image sequence, and select other original images in other original image sequences according to the certain position, wherein the other original image sequences are the original image sequences after removing a certain original image sequence from all original image sequences;

[0025] The second output module is configured to input the other original images into the image recognition model, and the image recognition model outputs a second recognition result corresponding to the other original images. Based on the second recognition result and the first recognition result, a preset impurity analysis strategy is used to perform basic nickel carbonate impurity analysis on the original image sequence and each original image in the other original image sequences to obtain the analysis results of basic nickel carbonate on each production line.

[0026] Thirdly, an electronic device is provided, comprising: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the steps of the basic nickel carbonate impurity analysis method according to any embodiment of the present invention.

[0027] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the program instructions are executed by a processor, the processor performs the steps of the basic nickel carbonate impurity analysis method according to any embodiment of the present invention.

[0028] The basic nickel carbonate impurity analysis method and system of this application constructs an adaptive threshold system through a dynamic slope clustering algorithm. In the early production stage, it can amplify weak signals and effectively identify initial basic nickel carbonate contamination. In the later stage, it automatically suppresses slowly changing interferences such as equipment wear. Compared with the fixed threshold method, the false alarm rate is reduced. Furthermore, it adopts a time-aligned clustering analysis strategy. When a certain serial number of a certain production line is identified as abnormal, it automatically and synchronously detects images of the same serial number on other production lines. Through shared equipment correlation analysis, the contamination source can be located relatively quickly. The basic nickel carbonate impurity analysis is performed on a certain original image sequence and other original image sequences using a preset impurity analysis strategy to obtain the analysis results of basic nickel carbonate on each production line. This can effectively improve the efficiency of basic nickel carbonate impurity analysis in the entire production line. Attached Figure Description

[0029] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0030] Figure 1 A flowchart of a method for analyzing impurities in basic nickel carbonate provided in an embodiment of the present invention;

[0031] Figure 2 This is a structural block diagram of a basic nickel carbonate impurity analysis system provided in an embodiment of the present invention;

[0032] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0033] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0034] Please see Figure 1 The diagram shows a flowchart of a method for analyzing impurities in basic nickel carbonate according to this application.

[0035] like Figure 1 As shown, the method for analyzing impurities in basic nickel carbonate specifically includes the following steps:

[0036] Step S101: Acquire raw images of basic nickel carbonate on each production line within a preset time period, and sort the raw images by time based on different production lines to obtain at least one raw image sequence.

[0037] In a specific application scenario, during the production of basic nickel carbonate, different production lines are needed to transport basic nickel carbonate to corresponding locations. However, during the transportation process, basic nickel carbonate may contain sudden impurities, such as metal fragments or other raw material impurities. Therefore, it is necessary to acquire the original images of basic nickel carbonate on each production line within a preset time period and execute subsequent steps S102-S106 to perform impurity analysis on the basic nickel carbonate.

[0038] Step S102: Obtain at least one image change rate of the original images in the same production line within a preset time period. Based on the at least one image change rate, use a preset image extraction strategy to extract images from each original image sequence to obtain a target original image set corresponding to each original image sequence.

[0039] In this step, the similarity and time interval between the second original image and the first original image in an original image sequence are obtained. The first original image is the first image in the original image sequence, and the second original image is any original image in the original image sequence excluding the first original image. The image change rate of the second original image is calculated based on the ratio of the similarity to the time interval. Specifically, the similarity between the second and first original images can be obtained by extracting high-level features (output of fully connected layers or pooling layers) using a pre-trained model (such as VGG or ResNet) and calculating cosine similarity or Euclidean distance.

[0040] It should be noted that, based on the sorting results of each original image in an original image sequence, at least one image change rate is sorted to obtain an image change rate sequence; a two-dimensional coordinate system is constructed, and each image change rate in the image change rate sequence is set in the two-dimensional coordinate system, and each image change rate is connected to the origin of the coordinate system to obtain at least one line segment, where the horizontal axis of the two-dimensional coordinate system is the sequence number corresponding to each image change rate in the image change rate sequence, and the vertical axis is the value of each image change rate in the image change rate sequence; the slope of at least one line segment is calculated, and line segments with slopes within a preset dynamic slope range are defined as normal line segments, and the rest are abnormal line segments, and the slopes corresponding to the abnormal line segments are defined as abnormal slopes, where the preset dynamic slope range is the slope range set with the slope cluster center as the range center; target original images corresponding to each abnormal slope are extracted from the original image sequence, and each target original image is divided into the same target original image set to obtain a target original image set corresponding to the original image sequence.

[0041] In a specific application scenario, the production line consists of three parallel production lines (A, B, and C) that produce basic nickel carbonate.

[0042] Data acquisition: Images of the basic nickel carbonate surface are acquired hourly for 24 consecutive hours (24 images per production line).

[0043] Image change rate: The image change rate is calculated by the pixel difference (such as SSIM difference) between the images at each acquisition time and the image at the first acquisition time, reflecting the state change of basic nickel carbonate.

[0044] The horizontal axis (X-axis) is the sequence number of the rate of change of the image (i.e., the number of the 1st rate of change, the 2nd rate of change, etc., starting from 1 and increasing).

[0045] Vertical axis (Y-axis): The numerical value of the rate of change of the image (such as the SSIM difference value).

[0046] Slope calculation: The slope of the line connecting each data point (xi,yi) to the origin (0,0) is ki = yi / xi. ki is the slope of the i-th line segment, yi is the ordinate of the i-th data point, and xi is the abscissa of the i-th data point.

[0047] Assuming that by clustering all slopes, the slope cluster center is 0.005, then the dynamic slope range is [0.005-0.01, 0.005+0.01].

[0048] Table 1

[0049] ,

[0050] Extract the original images corresponding to the abnormal slope (such as the original images of serial numbers 1, 2, 4, 5 and 24 in Table 1) into the target original image set. Then, if only the original image of serial number 24 is identified as abnormal in production line A, the original image of serial number 24 in production lines B / C is checked simultaneously.

[0051] In this embodiment, by using sequence number-based slope dynamic analysis and cross-production line timing alignment verification, the accuracy, sensitivity, and traceability of impurity detection in basic nickel carbonate production are significantly improved. The specific technical effects and comparative analysis are as follows:

[0052] In the slope calculation ki=yi / xi, the sequence number xi is used as the denominator. This amplifies early (small xi) abrupt changes (e.g., y2=0.05) (k2=0.025), while suppressing the same rate of change in later (large xi) stages (e.g., when y24=0.4, k24=0.0167). Minor anomalies (e.g., impurities in basic nickel carbonate) can be detected as early as sequence number 2 (early production stage), and slow changes later (e.g., equipment wear) will not be misjudged due to slope decay.

[0053] The slope is determined by data clustering, such as [0.01-0.01, 0.01+0.01], rather than a fixed threshold. The impact of production rhythm fluctuations (such as changes in feeding intervals) on detection is automatically eliminated.

[0054] Step S103: Input each target original image in a certain target original image set into a preset image recognition model, and the image recognition model outputs a first recognition result corresponding to each target original image.

[0055] In this step, within the impurity detection scheme for basic nickel carbonate production, the core task of the image recognition model is to accurately analyze the original target image corresponding to slope anomalies. The specific implementation process is as follows: First, the image is preprocessed (normalized, contrast enhanced) and multi-scale feature extracted using a convolutional neural network (such as EfficientNet-B3). Then, a classification head (fully connected layer + Softmax) or a detection head (such as the RPN network of Faster R-CNN) outputs the impurity category, location, and confidence level, generating structured results (such as JSON format). This image recognition model is trained using historical impurity image data and employs Focal Loss to address the sample imbalance problem. It achieves 96% accuracy and a real-time inference speed of 42ms at a 512×512 resolution, and can identify known impurities (such as metal particles and reactor debris).

[0056] Step S104: Determine whether each first identification result is an abnormal identification result, wherein the abnormal identification result is the identification result of impurities in basic nickel carbonate.

[0057] In one specific embodiment, if none of the first identification results are abnormal identification results, then each identification result is assigned to each original image in a certain original image sequence to obtain the analysis results of basic nickel carbonate on a certain production line. The certain original image sequence is the original image sequence obtained by sorting each original image from a certain production line by time, and the certain original image sequence corresponds to a certain target original image set. Each other target original image in the other target original image set is input into a preset image recognition model, and the image recognition model outputs a third identification result corresponding to each other target original image. It is then determined whether each third identification result is an abnormal identification result.

[0058] Step S105: If a certain first recognition result is an abnormal recognition result, then a certain target original image corresponding to the certain first recognition result is obtained, and a certain position of the certain target original image in a certain original image sequence is determined. Other original images are selected from other original image sequences according to the certain position, wherein the other original image sequences are the original image sequences after removing a certain original image sequence from all original image sequences.

[0059] In this step, a certain position of the target original image in a certain original image sequence is obtained, wherein the certain position is the sequence number of the target original image in the certain original image sequence; other original images with the same position are selected from other original image sequences.

[0060] In one specific embodiment, three basic nickel carbonate production lines, A, B, and C, share the same basic nickel carbonate conveying pipeline. The detection objective is to locate the source of impurities by analyzing the time-series images of each production line using an image recognition model.

[0061] Single production line anomaly detection, specifically:

[0062] Image Acquisition: Production line A acquires 24 images (sequence numbers 1-24) in chronological order, forming the original image sequence A. The image recognition model detects an anomaly (such as black particles, confidence level 92%) in image sequence number 18 (the target original image).

[0063] Location of the anomaly: The position of the anomaly image in sequence A is sequence number 18 (i.e., the image acquired at the 18th hour).

[0064] The cross-production line alignment analysis is as follows:

[0065] Select images with the same serial number from other production lines: Extract image number 18 from original image sequence B of production line B. Extract image number 18 from original image sequence C of production line C.

[0066] Synchronous identification results: Image 18 of production line B: Normal (no impurities, 98% confidence). Image 18 of production line C: Abnormal (same type of black particles, 90% confidence).

[0067] Conclusion: The abnormal pollution source location only occurred in serial number 18 of production lines A and C, while production line B was normal. Therefore, the pollution source is a shared equipment of lines A and C (such as a branch valve in the conveying pipeline).

[0068] If only line A is abnormal, then the specific batch of basic nickel carbonate for line A or the inner wall of the reactor needs to be checked.

[0069] Step S106: Input the other original images into the image recognition model. The image recognition model outputs a second recognition result corresponding to the other original images. Based on the second recognition result and the first recognition result, a preset impurity analysis strategy is used to perform basic nickel carbonate impurity analysis on the original image sequence and each original image in the other original image sequences to obtain the analysis results of basic nickel carbonate on each production line.

[0070] In this step, based on the first identification result, a preset impurity analysis strategy is used to perform basic nickel carbonate impurity analysis on each original image in a given original image sequence, obtaining the analysis results of basic nickel carbonate on each production line, specifically including:

[0071] Obtain the target original image corresponding to each first recognition result, and use each target original image as the cluster center in a certain original image sequence to divide at least one cluster set in a certain original image sequence; define each first recognition result as the recognition result of each original image in the corresponding cluster set, and obtain the analysis result of all original images in a certain original image sequence, wherein half of the original images between two adjacent target original images are divided into one cluster set, and the other half of the original images are divided into another cluster set.

[0072] Furthermore, based on the second identification result, a preset impurity analysis strategy is used to perform basic nickel carbonate impurity analysis on each original image in other original image sequences, obtaining the analysis results of basic nickel carbonate on each production line, specifically including:

[0073] Obtain other original images corresponding to each second recognition result, and use each other original image as the cluster center in the other original image sequence to divide at least one cluster set in the other original image sequence; define each second recognition result as the recognition result of each original image in the corresponding cluster set, and obtain the analysis results of all original images in the other original image sequence, wherein half of the original images between two adjacent other original images are divided into one cluster set, and the other half of the original images are divided into another cluster set.

[0074] In one specific embodiment, production lines A, B, and C share the same basic nickel carbonate conveying system. Image acquisition: Each production line acquires one image every hour, for a total of 24 images (serial numbers 1 to 24) over 24 hours.

[0075] Abnormal events:

[0076] Image 10 of production line A shows black particles (first identification result, confidence level 95%).

[0077] Image 10 of serial number from production line B identifies particles of the same type (second identification result, 90% confidence level).

[0078] The image of serial number 10 from production line C is normal (second identification result, confidence level 98%).

[0079] The cluster analysis for production line A is as follows:

[0080] Determine cluster centers:

[0081] Using the abnormal image (serial number 10) as the cluster center, divide the 24 images of production line A.

[0082] Clustering set partitioning rules:

[0083] Cluster set 1 (sequence numbers 1~10): contains sequence number 10 and its left-adjacent image.

[0084] Cluster set 2 (sequence numbers 11~24): the image adjacent to the right of sequence number 10.

[0085] Note: If there are multiple outliers (such as sequence numbers 5, 10, and 18), the division is based on the midpoint between adjacent outliers.

[0086] Analysis results:

[0087] Cluster set 1 (sequence numbers 1~10): marked as "containing black particles" (inheriting the first identification result of sequence number 10).

[0088] Cluster set 2 (sequence numbers 11~24): marked as "normal" (no abnormal identification results covered).

[0089] Since impurities are usually continuous over time (such as the persistence of basic nickel carbonate contamination), clustering can reduce the impact of misjudgment in a single frame.

[0090] Cluster analysis of production line B / C is as follows:

[0091] Production line B (serial number 10 abnormal):

[0092] Cluster set 1 (sequence numbers 1~10): marked as "containing black particles" (inheriting the second identification result of sequence number 10).

[0093] Cluster set 2 (sequence numbers 11~24): marked as "normal".

[0094] Production line C (serial number 10 is normal):

[0095] All images were labeled "normal" (no abnormal results drove clustering).

[0096] In summary, the method of this application constructs an adaptive threshold system through a dynamic slope clustering algorithm, which can amplify weak signals in the early production stage and effectively identify initial basic nickel carbonate contamination. In the later stage, it automatically suppresses slowly varying interferences such as equipment wear. Compared with the fixed threshold method, the false alarm rate is reduced. Furthermore, the method adopts a time-aligned clustering analysis strategy. When an anomaly is detected in a certain serial number of a certain production line, images of the same serial number on other production lines are automatically and synchronously detected. Through shared equipment correlation analysis, the contamination source can be located relatively quickly. The method uses a preset impurity analysis strategy to perform basic nickel carbonate impurity analysis on a certain original image sequence and other original image sequences to obtain the analysis results of basic nickel carbonate on each production line, which can effectively improve the efficiency of basic nickel carbonate impurity analysis in the entire production line.

[0097] Please see Figure 2 The diagram shows a structural block diagram of a basic nickel carbonate impurity analysis system according to this application.

[0098] like Figure 2 As shown, the basic nickel carbonate impurity analysis system 200 includes a sorting module 210, an extraction module 220, a first output module 230, a judgment module 240, a determination module 250, and a second output module 260.

[0099] The sorting module 210 is configured to acquire original images of basic nickel carbonate from various production lines within a preset time period, and sort the original images by time based on different production lines to obtain at least one original image sequence; the extraction module 220 is configured to acquire at least one image change rate of the original images from the same production line within a preset time period, and extract images from each original image sequence using a preset image extraction strategy based on the at least one image change rate to obtain a target original image set corresponding to each original image sequence; the first output module 230 is configured to input each target original image from a target original image set into a preset image recognition model, and the image recognition model outputs a first recognition result corresponding to each target original image; the judgment module 240 is configured to judge whether each first recognition result is an abnormal recognition result, wherein the abnormal recognition result is that the basic nickel carbonate contains impurities. The identification result determination module 250 is configured to, if a certain first identification result is an abnormal identification result, acquire a certain target original image corresponding to the certain first identification result, determine a certain position of the certain target original image in a certain original image sequence, and select other original images in other original image sequences according to the certain position, wherein the other original image sequences are the original image sequences after removing a certain original image sequence from all original image sequences; the second output module 260 is configured to input the other original images into the image recognition model, the image recognition model outputs a second identification result corresponding to the other original images, and, according to the second identification result and the first identification result, use a preset impurity analysis strategy to perform basic nickel carbonate impurity analysis on the certain original image sequence and each original image in the other original image sequences to obtain the analysis results of basic nickel carbonate on each production line.

[0100] It should be understood that Figure 2 The modules and references described in the document Figure 1 The steps described in the text correspond to those in the method described above. Therefore, the operations, features, and corresponding technical effects described above also apply to the method described in the text. Figure 2 The various modules in the document will not be described in detail here.

[0101] In other embodiments, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the program instructions are executed by a processor, the processor performs the basic nickel carbonate impurity analysis method in any of the above method embodiments.

[0102] In one embodiment, the computer-readable storage medium of the present invention stores computer-executable instructions, which are configured as follows:

[0103] The original images of basic nickel carbonate on each production line are acquired within a preset time period, and the original images are sorted by time based on different production lines to obtain at least one original image sequence.

[0104] Obtain at least one image change rate of the original images in the same production line within a preset time period. Based on the at least one image change rate, use a preset image extraction strategy to extract images from each original image sequence to obtain a target original image set corresponding to each original image sequence.

[0105] Each original target image in a set of original target images is input into a preset image recognition model, and the image recognition model outputs a first recognition result corresponding to each original target image;

[0106] Determine whether each first identification result is an abnormal identification result, wherein the abnormal identification result is the identification result of impurities in basic nickel carbonate;

[0107] If a certain first identification result is an abnormal identification result, then a certain target original image corresponding to the certain first identification result is obtained, and a certain position of the certain target original image in a certain original image sequence is determined. Other original images are selected from other original image sequences based on the certain position, wherein the other original image sequences are the original image sequences after removing a certain original image sequence from all original image sequences.

[0108] The other original images are input into the image recognition model, and the image recognition model outputs a second recognition result corresponding to the other original images. Based on the second recognition result and the first recognition result, a preset impurity analysis strategy is used to perform basic nickel carbonate impurity analysis on the original image sequence and each original image in the other original image sequences to obtain the analysis results of basic nickel carbonate on each production line.

[0109] Computer-readable storage media may include a stored program area and a stored data area, wherein the stored program area may store an operating system and an application program required for at least one function; the stored data area may store data created based on the use of the basic nickel carbonate impurity analysis system, etc. Furthermore, the computer-readable storage medium may include high-speed random access memory, and may also include memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, the computer-readable storage medium may optionally include memory remotely located relative to a processor, which can be connected to the basic nickel carbonate impurity analysis system via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0110] Figure 3 This is a schematic diagram of the structure of the electronic device provided in the embodiment of the present invention, such as... Figure 3 As shown, the device includes a processor 310 and a memory 320. The electronic device may also include an input device 330 and an output device 340. The processor 310, memory 320, input device 330, and output device 340 can be connected via a bus or other means. Figure 3 Taking a bus connection as an example, the memory 320 is the computer-readable storage medium described above. The processor 310 executes various server functions and data processing by running non-volatile software programs, instructions, and modules stored in the memory 320, thereby implementing the basic nickel carbonate impurity analysis method described in the above embodiment. The input device 330 can receive input digital or character information and generate key signal inputs related to user settings and function control of the basic nickel carbonate impurity analysis system. The output device 340 may include a display screen or other display device.

[0111] The aforementioned electronic device can execute the method provided in the embodiments of the present invention, and has the corresponding functional modules and beneficial effects for executing the method. Technical details not described in detail in this embodiment can be found in the method provided in the embodiments of the present invention.

[0112] In one embodiment, the above-described electronic device is applied in a basic nickel carbonate impurity analysis system for a client application, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to:

[0113] The original images of basic nickel carbonate on each production line are acquired within a preset time period, and the original images are sorted by time based on different production lines to obtain at least one original image sequence.

[0114] Obtain at least one image change rate of the original images in the same production line within a preset time period. Based on the at least one image change rate, use a preset image extraction strategy to extract images from each original image sequence to obtain a target original image set corresponding to each original image sequence.

[0115] Each original target image in a set of original target images is input into a preset image recognition model, and the image recognition model outputs a first recognition result corresponding to each original target image;

[0116] Determine whether each first identification result is an abnormal identification result, wherein the abnormal identification result is the identification result of impurities in basic nickel carbonate;

[0117] If a certain first identification result is an abnormal identification result, then a certain target original image corresponding to the certain first identification result is obtained, and a certain position of the certain target original image in a certain original image sequence is determined. Other original images are selected from other original image sequences based on the certain position, wherein the other original image sequences are the original image sequences after removing a certain original image sequence from all original image sequences.

[0118] The other original images are input into the image recognition model, and the image recognition model outputs a second recognition result corresponding to the other original images. Based on the second recognition result and the first recognition result, a preset impurity analysis strategy is used to perform basic nickel carbonate impurity analysis on the original image sequence and each original image in the other original image sequences to obtain the analysis results of basic nickel carbonate on each production line.

[0119] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of embodiments.

[0120] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for the impurity analysis of basic nickel carbonate, characterized in that, The method comprises the following steps: acquiring original images of basic nickel carbonate on each production line within a preset time period, and performing time sorting on each original image based on different production lines to obtain at least one original image sequence; acquiring at least one image change rate of the original images in the same production line within the preset time period, and performing image extraction on each original image sequence based on a preset image extraction strategy according to the at least one image change rate to obtain a target original image set corresponding to each original image sequence; inputting each target original image in a certain target original image set into a preset image recognition model, and obtaining a first recognition result corresponding to each target original image through the image recognition model; judging whether each first recognition result is an abnormal recognition result, wherein the abnormal recognition result is a recognition result of basic nickel carbonate containing impurities; if a certain first recognition result is an abnormal recognition result, acquiring a certain target original image corresponding to the certain first recognition result, determining a certain position of the certain target original image in a certain original image sequence, and selecting other original images in other original image sequences according to the certain position, wherein the other original image sequences are original image sequences after removing the certain original image sequence from all original image sequences, and the determining of the certain position of the certain target original image in the certain original image sequence and the selecting of the other original images in the other original image sequences comprise: acquiring a certain position of the certain target original image in the certain original image sequence, wherein the certain position is a sequence number of the certain target original image in the certain original image sequence; selecting other original images with the same certain position in the other original image sequences; inputting the other original images into the image recognition model, obtaining a second recognition result corresponding to the other original images through the image recognition model, and performing basic nickel carbonate impurity analysis on each original image in the certain original image sequence and the other original image sequences according to a preset impurity analysis strategy based on the second recognition result and the first recognition result to obtain an analysis result of basic nickel carbonate on each production line, wherein the performing of the basic nickel carbonate impurity analysis on each original image in the certain original image sequence according to the first recognition result and the preset impurity analysis strategy to obtain the analysis result of basic nickel carbonate on each production line comprises: acquiring target original images corresponding to each first recognition result, taking each target original image as a clustering center in a certain original image sequence, and dividing at least one cluster set in the certain original image sequence; defining each first recognition result as a recognition result of each original image in the corresponding cluster set to obtain an analysis result of all original images in the certain original image sequence, wherein half of the original images between two adjacent target original images are divided into one cluster set, and the other half of the original images are divided into another cluster set.

2. The method for impurity analysis of basic nickel carbonate according to claim 1, characterized in that, The acquiring of the at least one image change rate of the original images in the same production line within a preset time period comprises: acquiring a similarity and a time interval between a second original image and a first original image in a sequence of original images, wherein the first original image is an original image ranked first in the sequence of original images, and the second original image is any original image in the sequence of original images except the first original image; calculating an image change rate of the second original image according to a ratio of the similarity to the time interval.

3. The method for impurity analysis of basic nickel carbonate according to claim 1, characterized in that, The image extraction of each sequence of original images according to the at least one image change rate and a preset image extraction strategy to obtain a target original image set corresponding to each sequence of original images comprises: sorting the at least one image change rate according to a ranking result of each original image in the sequence of original images to obtain a sequence of image change rates; constructing a two-dimensional coordinate system, setting each image change rate in the sequence of image change rates in the two-dimensional coordinate system, and connecting each image change rate with the coordinate origin to obtain at least one line segment, wherein the horizontal coordinate of the two-dimensional coordinate system is a sequence number corresponding to each image change rate in the sequence of image change rates, and the vertical coordinate is a value of each image change rate in the sequence of image change rates; calculating a slope of the at least one line segment, defining a line segment with a slope within a preset dynamic slope range as a normal line segment, and defining the rest as an abnormal line segment, and defining a slope corresponding to the abnormal line segment as an abnormal slope, wherein the preset dynamic slope range is a slope range with a slope clustering center as a range center; extracting a target original image corresponding to each abnormal slope in the sequence of original images, and dividing each target original image into the same target original image set to obtain a target original image set corresponding to the sequence of original images.

4. The method for impurity analysis of basic nickel carbonate according to claim 1, characterized in that, After judging whether each first recognition result is an abnormal recognition result, the method further comprises: if none of the first recognition results is an abnormal recognition result, distributing the first recognition results to each original image in the sequence of original images to obtain an analysis result of basic nickel carbonate on the production line, wherein the sequence of original images is a sequence of original images obtained by time sorting of the original images from the production line, and the sequence of original images corresponds to the target original image set; inputting each other target original image in the other target original image set into a preset image recognition model, and outputting a third recognition result corresponding to each other target original image from the image recognition model; judging whether each third recognition result is an abnormal recognition result.

5. The method for impurity analysis of basic nickel carbonate according to claim 1, characterized by, wherein performing basic nickel carbonate impurity analysis on each original image in the other sequence of original images according to the second recognition result and a preset impurity analysis strategy to obtain an analysis result of basic nickel carbonate on each production line, specifically comprising: obtain other original images corresponding to each second recognition result, take each other original image as a clustering center in an other original image sequence, and divide at least one clustering set in the other original image sequence; define each second recognition result as a recognition result of each original image in a corresponding clustering set, and obtain an analysis result of all original images in the other original image sequence, wherein, one half of original images between two adjacent other original images are divided into one clustering set, and the other half of original images are divided into another clustering set.

6. A system for impurity analysis of basic nickel carbonate, characterized by comprise: an ordering module configured to obtain original images of basic nickel carbonate on each production line within a preset time period, and perform time ordering on each original image based on different production lines to obtain at least one original image sequence; an extraction module configured to obtain at least one image change rate of original images in a same production line within a preset time period, and perform image extraction on each original image sequence based on a preset image extraction strategy according to the at least one image change rate to obtain a target original image set corresponding to the each original image sequence; a first output module configured to input each target original image in a certain target original image set into a preset image recognition model, and output a first recognition result corresponding to the each target original image from the image recognition model; a judgment module configured to judge whether each first recognition result is an abnormal recognition result, wherein the abnormal recognition result is a recognition result of basic nickel carbonate containing impurities; a determination module configured to, if a certain first recognition result is an abnormal recognition result, obtain a certain target original image corresponding to the certain first recognition result, determine a certain position of the certain target original image in a certain original image sequence, and select other original images in other original image sequences according to the certain position, wherein the other original image sequences are original image sequences after removing the certain original image sequence from all original image sequences, and the determination of the certain position of the certain target original image in the certain original image sequence and the selection of the other original images in the other original image sequences comprise: obtaining a certain position of the certain target original image in the certain original image sequence, wherein the certain position is a sequence number of the certain target original image in the certain original image sequence; selecting other original images with the same certain position in the other original image sequences; The second output module is configured to input the other original images into the image recognition model, and the image recognition model outputs a second recognition result corresponding to the other original images. According to the second recognition result and the first recognition result, a preset impurity analysis strategy is used to analyze the alkali carbonate nickel impurities in each original image in the certain original image sequence and the other original image sequence, and an analysis result of the alkali carbonate nickel on each production line is obtained. According to the first recognition result, the preset impurity analysis strategy is used to analyze the alkali carbonate nickel impurities in each original image in the certain original image sequence, and the analysis result of the alkali carbonate nickel on each production line is obtained. Specifically, the method comprises the following steps: Obtaining target original images corresponding to each first recognition result, taking each target original image as a clustering center in a certain original image sequence, and dividing at least one clustering set in the certain original image sequence; Defining each first recognition result as a recognition result of each original image in the corresponding clustering set, and obtaining an analysis result of all original images in the certain original image sequence, wherein half of the original images between two adjacent target original images are divided into one clustering set, and the other half of the original images are divided into another clustering set.

7. An electronic device, comprising: Comprise: At least one processor and a memory connected in communication with the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1 to 5.

8. A computer-readable storage medium having stored thereon a computer program, characterized in that The program is executed by the processor to implement the method of any one of claims 1 to 5.

Images