Search

CN-121999306-A - Color screen instrument image classification method and system based on RLE decompression

CN121999306ACN 121999306 ACN121999306 ACN 121999306ACN-121999306-A

Abstract

The invention relates to the technical field of image processing and provides a color screen instrument image classification method and system based on RLE decompression, wherein the method comprises the steps of acquiring vehicle running state data, and performing first verification on a first recognition result based on the vehicle running state data to obtain a first verification conclusion; carrying out second identification on a local area containing safety key information in the data stream to obtain a second identification conclusion; if the safety key information category indicated by the first identification result is consistent with the safety key information categories indicated by the first verification conclusion and the second identification conclusion, the first identification result is displayed with the highest priority, and if the safety key information category indicated by the first identification result is inconsistent with the safety key information categories indicated by the first verification conclusion and the second identification conclusion, the first identification result is not adopted and the rollback operation is executed. The method has the effect of improving the accuracy and reliability of the color screen instrument image classification.

Inventors

  • SUN QUANJIN

Assignees

  • 深圳市迪太科技有限公司

Dates

Publication Date
20260508
Application Date
20260410

Claims (10)

  1. 1. The color screen instrument image classification method based on RLE decompression is characterized by comprising the following steps of: In the process of decompressing the data stream of the RLE compressed image, carrying out preliminary identification on the data stream according to a preset safety key RLE mode signature, wherein the preset safety key RLE mode signature is a signature template for representing the RLE stroke sequence form of a safety key display area, and the signature template at least comprises a key area identifier, a key color category identifier, a stroke length sequence feature and a matching threshold configuration; When safety key information is identified in the primary identification process, a first identification result is output and a preset high-priority verification process is started, wherein the high-priority verification process comprises the steps of sending a diagnosis request to an external control unit, acquiring vehicle running state data corresponding to the safety key information, and carrying out first verification on the first identification result based on the vehicle running state data to obtain a first verification conclusion; And finally confirming based on the first identification result, the first verification conclusion and the second identification conclusion, if the safety key information category indicated by the first identification result is consistent with the safety key information category indicated by the first verification conclusion and the second identification conclusion, displaying the first identification result with the highest priority, and if the safety key information category indicated by the first identification result is inconsistent with the safety key information categories indicated by the first verification conclusion and the second identification conclusion, not adopting the first identification result and executing rollback operation.
  2. 2. The method for classifying color screen instrument images based on RLE decompression according to claim 1, wherein the step of primarily identifying the data stream of the RLE compressed image according to a preset security-critical RLE mode signature during the decompression process of the data stream comprises: The method comprises the steps of carrying out rule aggregation on data streams of RLE compressed images, starting rule aggregation judgment when the rule aggregation is started, if a section of expected dominant hue is detected to be interrupted by a run segment in a window, wherein the window is a preset decompression scanning range, the run is a length segment of continuous same-color pixels in the RLE, the run segment is an interrupted or split run sub-segment, judging whether the run length of the interrupted segment is lower than a run length threshold value after the rule aggregation judgment is started to obtain a run length judgment result, judging whether the hue difference and the saturation difference of the interrupted segment color and the expected dominant hue in a color space are smaller than a difference threshold value to obtain a color difference judgment result, merging or rejecting the interrupted segment according to preset rules when the run length judgment result and the color difference judgment result are both represented as yes, and reconstructing the interrupted segment into a continuous run; And matching the continuous travel obtained based on rule aggregation with a preset safety critical RLE mode signature so as to primarily identify the data flow.
  3. 3. The RLE decompression based color screen instrument image classification method according to claim 1, wherein the step of performing a first verification on the first recognition result based on the vehicle running state data to obtain a first verification conclusion includes: sending a diagnosis request to an external control unit corresponding to the safety key information; after the diagnosis request is sent, capturing the response of the external control unit and analyzing the response state corresponding to the diagnosis request, wherein the response state comprises clear fault, clear no fault, busy system, old data or no response; continuously collecting vehicle running state data, wherein the vehicle running state data comprises vehicle speed, engine load and environment temperature, and maintaining a real-time updated vehicle state parameter table; The method comprises the steps of maintaining an event log in a preset time period, storing a first identification result, a response state of an external control unit and key operation parameter fluctuation in the last time period, and identifying repeated, continuous or aggravated abnormal modes; When the response state is that the system is busy, old data or no response is generated, calculating a current security risk probability score based on a first identification result, the response state, vehicle running state data and an event log, and adopting a preset self-adaptive warning display strategy according to the security risk probability score as a first verification conclusion; When the response state is an explicit fault, displaying a determined safety warning with the highest priority as a first verification conclusion; And when the response state is clear and fault-free, marking the first identification result as a low-risk false alarm candidate, executing degradation display or non-display processing, and recording the corresponding event into an event log to serve as a subsequent statistical calibration basis to serve as a first verification conclusion.
  4. 4. The method for classifying color screen instrument images based on RLE decompression according to claim 1, wherein the step of performing a second recognition on the local area containing the safety key information in the data stream according to the relative color relationship and the structural characteristics between pixels in the local area, and obtaining a second recognition conclusion includes: The background separation process comprises the steps of identifying a preset color area which is related to the safety key information in the local area, and regularly distinguishing the preset color area from surrounding background areas according to the boundary of the preset color area, wherein the rule distinction is to divide a target area from the background area according to the preset color area boundary; based on the rule distinguishing result, calculating the relative brightness difference and the relative color difference between pixels in the safety key information area, and analyzing the travel length distribution in the safety key information area; calculating the color contrast, the boundary definition and the structure complexity difference between the safety key information area and the background area according to the relative brightness difference and the relative color difference between pixels in the safety key information area and the stroke length distribution; and matching the difference of the color contrast, the boundary definition and the structural complexity with a preset safety key information template to obtain a second identification conclusion.
  5. 5. The RLE decompression based color screen instrument image classification method according to claim 3, wherein the step of calculating a current security risk likelihood score based on the first identification result, the response status, the vehicle operating status data, and the event log comprises: the method comprises the steps of obtaining the current driving working condition of a vehicle, wherein the driving working condition comprises the speed of the vehicle, the acceleration of the vehicle, the braking state and the steering angle; according to the driving working condition, adjusting the first identification result, the response state of the external control unit, the weight of the vehicle running state data and the event log in scoring calculation; Adjusting a risk threshold according to the driving condition; And calculating the current safety risk probability score according to the adjusted weight and the risk threshold value and based on the first identification result, the response state, the vehicle running state data and the event log.
  6. 6. The RLE decompression based color screen instrument image classification method according to claim 3, wherein the step of calculating a current security risk likelihood score based on the first identification result, the response status, the vehicle operating status data, and the event log comprises: Reading a first identification result, a response state of an external control unit, vehicle running state data and an event log; The dynamic degree of the driving working condition is estimated according to the speed change rate, the steering angle change rate and the braking strength of the vehicle, wherein the dynamic degree is a quantitative index for representing the intensity of the driving working condition change in a preset time window, and the dynamic degree is at least determined by one or more of the speed change rate, the acceleration fluctuation amplitude or the steering angular speed fluctuation amplitude and is used for triggering the linkage adjustment of the timeliness weight, the accuracy weight and the stability threshold; Based on the dynamic degree, timeliness weight and accuracy weight are distributed to the first identification result, the response state, the vehicle running state data and the event log; taking the first identification result, the response state, the vehicle running state data and the event log as various different information sources; Performing time stamp calibration and data integrity check on each information source to obtain a time stamp calibration result and a data integrity check result; The method comprises the steps of obtaining a time stamp calibration result, namely, judging whether the time stamp calibration result is lower than a preset threshold value or not, judging whether a data integrity check result is lower than a data integrity check result, and judging whether the time stamp calibration result is lower than a data integrity check result or not, wherein the time stamp calibration result is lower than a preset threshold value or not; and calculating the current security risk possibility score according to the timeliness weight, the accuracy weight and the calibrated various information sources.
  7. 7. The RLE decompression based color screen instrument image classification method according to claim 6, wherein the step of assigning timeliness and accuracy weights to the first recognition result, the response status, the vehicle operation status data, and the event log based on the dynamic level comprises: acquiring the speed change rate, the steering angle change rate and the braking strength of the vehicle; Calculating trend intensity indexes of the speed change rate, the steering angle change rate and the braking intensity; Triggering the pre-switching of the driving condition dynamic degree when the trend intensity index exceeds a preset dynamic switching threshold value, wherein the pre-switching comprises the steps of improving the timeliness weight and the accuracy weight related to the high dynamic condition and reducing the timeliness weight and the accuracy weight related to the stable driving condition; and when the trend intensity index is continuously lower than the dynamic fallback threshold value, reducing the timeliness weight and the accuracy weight related to the high dynamic working condition, and recovering the timeliness weight and the accuracy weight related to the stable driving working condition.
  8. 8. The RLE decompression based color screen instrument image classification method according to claim 3, wherein the step of calculating a current security risk likelihood score based on the first identification result, the response status, the vehicle operating status data, and the event log comprises: Acquiring a current driving mode of a vehicle; According to the driving mode, adjusting the weight of the first identification result, the response state of the external control unit, the vehicle running state data and the event log in scoring calculation; And calculating the current safety risk possibility score based on the first identification result, the response state, the vehicle running state data and the event log and according to the adjusted weight.
  9. 9. The RLE decompression based color screen instrument image classification method according to claim 3, wherein the step of calculating a current security risk likelihood score based on the first identification result, the response status, the vehicle operating status data, and the event log comprises: Acquiring the current running state of the vehicle; judging whether the running state indicates that the vehicle is in a state lower than a preset speed or in a parking state; When the vehicle is in a state lower than a preset speed or in a parking state, sending a detection request to an external control unit to acquire a response state corresponding to the detection request, and replacing the response state corresponding to the diagnosis request with the response state corresponding to the detection request; continuously reading a conventional running state message from an external control unit on a vehicle-mounted communication bus, and analyzing the data updating frequency and the numerical fluctuation of the conventional running state message; Based on the data update frequency and the numerical fluctuation, identifying whether there is a change in other vehicle operation state data; Marking the current state as an expected uncertainty under low dynamic conditions when the response state indicates that the external control unit continues to return an uncertain response and other vehicle operating state data is unchanged; When the expected uncertainty under the low-dynamic working condition appears, the negative influence of the uncertainty response returned by the external control unit on the safety risk possibility score is reduced, and the weight of other vehicle running state data in the score is improved; The method comprises the steps of determining the duration of uncertain response of an external control unit, and increasing a stability threshold according to the duration of uncertain response, wherein the stability threshold is a tolerance duration threshold for determining uncertain response of the external control unit under a low dynamic working condition, and the stability threshold is increased and used for delaying triggering of negative risk accumulation so as to avoid misjudgment of expected uncertainty caused by parking or low speed as high risk; and calculating the current safety risk possibility score based on the first identification result, the response state, the vehicle running state data and the event log and according to the adjusted weight and the stability threshold value.
  10. 10. A RLE decompression based color screen instrument image classification system for performing RLE decompression based color screen instrument image classification, comprising: The device comprises a primary identification execution module, a primary identification execution module and a primary identification module, wherein the primary identification execution module is used for carrying out primary identification on a data stream of an RLE compressed image according to a preset safety key RLE mode signature in the process of carrying out decompression processing on the data stream, wherein the preset safety key RLE mode signature is a signature template for representing the RLE stroke sequence form of a safety key display area, and the signature template at least comprises a key area identifier, a key color category identifier, a stroke length sequence feature and a matching threshold configuration; The identification verification conclusion module is used for outputting a first identification result and starting a preset high-priority verification process when safety key information is identified in the primary identification process; the high-priority verification process comprises the steps of sending a diagnosis request to an external control unit to acquire vehicle running state data corresponding to the safety key information, and carrying out first verification on the first recognition result based on the vehicle running state data to obtain a first verification conclusion; And the recognition result processing module is used for carrying out final confirmation based on the first recognition result, the first verification conclusion and the second recognition conclusion, if the safety key information category indicated by the first recognition result is consistent with the safety key information category indicated by the first verification conclusion and the second recognition conclusion, carrying out display processing on the first recognition result with the highest priority, and if the safety key information category indicated by the first recognition result is inconsistent with the safety key information categories indicated by the first verification conclusion and the second recognition conclusion, not adopting the first recognition result and executing rollback operation.

Description

Color screen instrument image classification method and system based on RLE decompression Technical Field The invention relates to the technical field of image processing, in particular to a color screen instrument image classification method and system based on RLE decompression. Background In modern intelligent driving vehicles, the color screen instrument system is a core window for a driver to acquire vehicle information, and is also a key interface for realizing intelligent driving and vehicle-mounted interaction. In order to efficiently utilize the limited memory space and ensure the smoothness of data transmission, instrumentation systems often employ Run Length Encoding (RLE) and the like technology to store various types of image data, such as navigation maps, multimedia interfaces, vehicle status displays, and various warning icons. When the system needs to present these compressed images to the driver, RLE decompression is required to restore the data to pixel information for display. However, merely decompressing the image is insufficient, and the system needs to quickly and accurately recognize the decompressed image contents to decide how to perform subsequent display processing, such as overlaying a route on a map or displaying a real-time vehicle speed on a dashboard, according to its category. In the prior art, this identification of image content is typically done by a separate classification module after the RLE decompression is completed. The serial processing mode of decompression before classification often causes obvious system delay when facing the increasing demands of intelligent driving on the real-time property and the richness of display contents, thereby influencing the operation experience and the driving safety of drivers. In order to improve the response speed of the system, an improved scheme is proposed, which performs image classification synchronously in the process of performing RLE decompression. The decompression and classification integrated processing mode omits the intermediate step of generating a complete bitmap for classification and corresponding memory occupation, and can theoretically remarkably reduce display delay. However, in practical vehicle applications, this integrated solution may face serious reliability problems. The graphic resource files of the in-vehicle system may be made at different times by different design teams or provided by different suppliers, which results in inconsistencies in the resource files themselves. For example, a warning icon with too low tire pressure, the design specification requires that its background color be a specific amber color, but there may be multiple versions in the actually stored resource file, where the background color of one version has a slight deviation from the standard value due to the derived setting error. For the integrated classification method for directly analyzing the RLE data stream, because the judgment basis may include accurate matching of specific color values, the deviation of the color values may cause the classifier to fail to recognize that the color values are a warning icon, so that a missing report occurs, and potential safety hazards are formed. A more serious problem than color deviation is data corruption. In-vehicle memories may experience occasional bit errors due to hardware aging or electromagnetic interference during long term use. RLE data formats are very sensitive to such errors, and flipping of one bit in the data may change a shorter run-length value to a very large value, or one color value to another, completely different color. When the decompressor processes this corrupted piece of data, a long, errant line may be drawn on the screen. The integrated classifier extracts the completely erroneous feature when analyzing this contaminated data stream, and most likely recognizes a red warning icon, which should be "engine failure", as a "multimedia play" interface, erroneously, thus presenting the user with completely erroneous information, which is absolutely unacceptable during driving. In view of the above, there is a need in the art for improvements. Disclosure of Invention The application discloses a color screen instrument image classification method and system based on RLE decompression, and aims to solve the problems of system delay and reliability in RLE decompression image classification and classification errors and potential safety hazards caused by inconsistent resource files or data damage in the prior art. The technical scheme of the application is as follows: In a first aspect, the application discloses a color screen instrument image classification method based on RLE decompression, which comprises the following steps: In the process of decompressing the data stream of the RLE compressed image, carrying out preliminary identification on the data stream according to a preset safety key RLE mode signature, wherein the preset safety key RLE mode signature is a