JP5962898B2 - Driving evaluation system, driving evaluation method, and driving evaluation program - Google Patents

Description

  The present invention relates to a driving evaluation system that collects driving records of a vehicle accompanied with a video and allows a driving evaluator to visually check the driving records, and more particularly to a driving evaluation system that can improve the efficiency of driving evaluation work.

  In recent years, driving evaluations of automobiles, motorcycles, bicycles, and the like have been performed in various environments.

  For driving evaluation, there are an evaluation method in which drive records are collected from a drive recorder, etc., and an evaluator visually determines whether the collected records are good or bad, and an evaluation method in which mechanical evaluation is performed from data such as fuel consumption and brake timing. .

Examples of techniques related to driving evaluation include Patent Documents 1 to 4.
Patent Document 1 describes a driving evaluation system including an in-vehicle device and an information processing center as shown in FIG. This driving evaluation system is a system that mechanically evaluates driving based on the collected information. In addition, the driving evaluation system calculates the places (checkpoints) that should be taken care of in the future driving and improves driving. Is configured to prompt.
More specifically, the in-vehicle device mounted on the automobile is vehicle information (accelerator opening, engine speed, vehicle speed, acceleration, fuel consumption, etc.) acquired from various in-vehicle sensors and EUC (Electronic Control Unit). Using the navigation information (travel route, road shape, road gradient, etc.) acquired from the car navigation system, the driver's driving in a predetermined traveling section where the vehicle traveled is evaluated and converted into data. Checkpoints are calculated so as to prompt the driver to improve driving, and an advice display is notified to the driver regarding the checkpoints. Further, it is described that each information processing may be performed on the information processing center side.

  Cited Document 2 is an invention by the applicant, and is an image obtained by acquiring a moving image from imaging means mounted on the own vehicle and extracting road environment information (road signs, road signs, signal states, pedestrians, etc.). Describes a system that performs recognition processing and notifies the higher-level server, as well as accepting road environment information acquired by other vehicles from the higher-level server and presenting the driver with road environment information that includes areas not captured by the imaging means in real time. doing. With this system, traffic rules that have not been fully computerized from road signs reflected in moving images can be computerized and presented to the driver in real time, including road environment information acquired by other vehicles.

  Patent Document 3 also describes an example of an image recognition technique for extracting a stop line that once causes a stop obligation.

  Patent Document 4 describes a driving evaluation device that performs driving evaluation without using a moving image, as in Patent Document 1. In this driving evaluation device, the driving (running position) of the person who is good at driving and the person to be evaluated is presented for comparison, and points to be corrected are transmitted to the person to be evaluated.

Japanese Unexamined Patent Publication No. 2011-081743 Japanese Patent No. 4752836 Japanese Patent Application No. 2009-099125 JP 2006-243248 A

  It is important for the driver's driving evaluation to reflect road signs, road signs, signal conditions, and the like. Such driving evaluation has been performed including traffic rules that have not been fully computerized together with the vehicle driver by riding or following the vehicle.

  In recent years, a method has also been used in which a driving moving image is recorded using a drive recorder, and driving evaluation is performed by viewing the moving image.

  The mechanical driving evaluation method described in the cited document 1 can be expected to greatly improve compliance with the legal speed and fuel efficiency. On the other hand, there are many points to be improved. For example, in the driving evaluation system of cited document 1, evaluation is based on vehicle information and navigation information, and road environment information detected from moving images in cited document 2 is evaluated. Cannot be reflected in In this case, for example, when the intersection intrusion at the yellow light or the ignorance of the stop line is temporarily performed as in the evaluation of the fuel consumption, a better result can be obtained as an evaluation result than in compliance with the traffic rules. Yes.

  In this way, driving evaluation obtained by visual observation of a moving image or the like can be an effective means for optimizing driving as compared with driving evaluation based on information already known.

  However, there is a problem in driving evaluation of the driver based on such riding, follow-up, and recorded moving images. For example, the problem that the evaluation time of the same time with respect to driving time is mentioned.

  In addition, driving evaluation using moving images or the like has a problem in quantifying individual driving evaluation, and there is a current situation in which a difference occurs in evaluation for each evaluator.

  In addition, a system that can suppress various costs for driving evaluation and can continuously perform driving evaluation on a daily basis is desired without performing limitedly as instructed by an instructor. In addition, a system that can continuously perform driving evaluation of many drivers is desired.

  There is also a demand to keep the in-vehicle device inexpensive when operating a large number of units.

  In addition, there is a demand to keep the operation cost for data transmission to the host server low. For example, there is a request to suppress the intensive generation of communication traffic and processing load.

  The present invention has been made in view of the above-described matters, and an object of the present invention is to provide a driving evaluation system capable of operating driving evaluation performed with reference to a moving image at low cost.

A driving evaluation system according to the present invention includes a moving image storage means for storing a moving image recorded by an imaging means mounted on a vehicle, and a reference object used for driving evaluation from the moving image recorded in the moving image storage means. The image recognition means for extracting and outputting at least the extraction position, the log storage means for storing the extraction position output by the image recognition means as an evaluation log used for driving evaluation, and the moving image storage means When evaluating a moving image, based on at least the extraction position of the reference object stored in the log storage means, the reproduction time of other sections is shortened while leaving a section in which the reference object included in the moving image appears. , that by shortening the regeneration processing time of the entire moving image to be evaluated and a log visualization means for reproducing processes applicable moving picture, may perform driving evaluation of the recorded moving image visually And butterflies.
The driving evaluation method according to the present invention includes a step of storing a moving image recorded by an imaging unit mounted on a vehicle, a reference object used for driving evaluation is extracted from the recorded moving image by image recognition processing, and at least Storing the extraction position as an evaluation log used for driving evaluation and when evaluating the recorded moving image, the moving image is included based on at least the extraction position of the reference object recorded in the evaluation log. A process of reducing the playback time of the entire moving image to be evaluated by reducing the playback time of other sections while leaving the section in which the reference object appears in the information processing system. The operation evaluation of the recorded moving image is performed visually.
The driving evaluation program according to the present invention includes a moving image storage unit that stores a moving image recorded by an imaging unit mounted on a vehicle, and a moving image recorded in the moving image storage unit. Extracting a reference object used for driving evaluation from the image recognition means for outputting at least the extraction position, log storage means for storing the extraction position output by the image recognition means as an evaluation log used for driving evaluation, When evaluating the moving image recorded in the moving image storage means, based on at least the extraction position of the reference object stored in the log storage means, while leaving the interval in which the reference object included in the moving image appears, other intervals Log visualization means for shortening the playback time of the entire moving image to be evaluated and reducing the playback processing time of the corresponding moving image,
As described above, the operation is performed so that the operation evaluation of the recorded moving image can be performed visually.

  ADVANTAGE OF THE INVENTION According to this invention, the driving | operation evaluation system which can operate | operate the driving | running evaluation performed with reference to a moving image at low cost can be provided.

It is a block diagram which shows the function structure of the driving | operation evaluation system 1 which concerns on 1st Embodiment. It is a block diagram which shows the structural example of a vehicle-mounted apparatus. 4 is a flowchart illustrating an image recognition processing flow illustrating a method for identifying a paused position in the image recognition processing unit 30. It is explanatory drawing which shows the example of a screen display output from the log visualization part. It is a block diagram which shows the function structure of the driving | operation evaluation system 2 which concerns on 2nd Embodiment. It is a block diagram which shows the function structure of the driving | operation evaluation system 3 which concerns on 3rd Embodiment. It is explanatory drawing which shows the driving | operation evaluation system which concerns on a 1st Example. It is explanatory drawing which shows the driving | operation evaluation system which concerns on a 2nd Example. It is explanatory drawing which shows the driving | operation evaluation system which concerns on a 3rd Example. It is explanatory drawing which shows the driving | operation evaluation system which concerns on a 4th Example. It is a block diagram which shows the driving | running evaluation system comprised by the vehicle-mounted apparatus shown by patent document 1, and information processing center.

An embodiment of the present invention will be described with reference to FIGS.
In the present embodiment, a driving evaluation system configured to collect moving images obtained from in-vehicle devices mounted on a large number of vehicles and perform visual safety evaluation for all of them will be described.

In this driving evaluation system, the driver of each vehicle passes a moving image taken during driving of the vehicle as data to the evaluator when receiving the evaluation, and individually receives driving evaluation.
Also, the evaluator is assigned an evaluator, and the driving evaluation for each individual is efficiently performed based on a large number of transferred video images.

  In the driving evaluation, whether or not the traffic rules based on the reference objects (stop lines, signs, etc.) used for the evaluation are being observed is mainly checked visually. In addition, as an evaluation item, a driving operation (vehicle speed during travel, how to step on a brake, etc.) with respect to the presence of a reference object is checked.

FIG. 1 is a block diagram illustrating a functional configuration of a driving evaluation system 1 according to an embodiment.
In FIG. 1, the present embodiment includes a moving image storage unit 10, a log storage unit 20, an image recognition processing unit 30, and a log visualization unit 40.

  The moving image storage unit 10 receives and stores the recorded moving image data. Any means may be used as the receiving method. For example, the image data may be received via a semiconductor storage medium or via a communication network such as a wireless LAN.

  The log storage unit 20 stores a log used for driving evaluation output from the image recognition processing unit 30 as an evaluation log. This evaluation log is associated with moving image data stored in the moving image storage unit 10.

  The image recognition processing unit 30 analyzes the moving image data stored in the moving image storage unit 10 by image recognition processing by an image recognition engine, and extracts a reference object used for driving evaluation.

  As the reference object, road signs, road signs, signal states, pedestrians, and the like are appropriately extracted in accordance with visual evaluation items. During this extraction process, the reference target extraction position is recorded in the log storage unit 20. The extraction position of the reference object may be recorded as a temporal position in the moving image, a spatial position in the moving image, a position in the real space, etc. as necessary. In addition to image identification such as signs, stop lines, and pedestrian crossings, ranges (positions, coordinates) that cause legal rules such as 5m and 10m from intersections, and ranges that apply private rules such as ○○ m from intersections and schools Or the like may be registered as an evaluation log as an extraction position.

When the log visualization unit 40 presents the moving image stored in the moving image storage unit 10 to the evaluator, the reference object included in the moving image is determined using the evaluation log stored in the log storage unit 20. With emphasis on the appearance timing, the content of the moving image is reproduced while reducing the reproduction processing time. The evaluator visually confirms the reproduced video to evaluate the driving of each individual driver.
In the evaluation log, as described above, the extraction position of the reference object according to the evaluation item to be performed is recorded. The reference object is a standard that defines a range, a place, a section, and an operation to be evaluated with priority. From the recording position (viewable position or position of the reference object) in a moving image where there is an individual reference object, from the timing before a predetermined time or a predetermined distance, for example, the playback of the moving image is set to the standard playback speed and other ranges Fast-forwarding and skipping are performed as appropriate, and the time is advanced faster than, for example, playback at the standard playback speed.
As a result, it is possible to maintain the evaluation accuracy and shorten the reproduction processing time for the visual evaluation at the standard reproduction speed and the visual evaluation by reproducing all at the double speed.

  Each of the above-described components can be operated by arranging a part thereof in a higher-level server. In this case, communication means may be provided as appropriate to connect the lower apparatus and the upper server via the Internet or a mobile communication network.

FIG. 2 is a block diagram illustrating a configuration of the in-vehicle device.
The in-vehicle device 100 includes an imaging unit 101, an operation data collection unit 102, a recording unit 103, and a storage unit 104. The in-vehicle device 100 is a so-called drive recorder, and the recording unit 103 converts video and sound in front of the vehicle into data and forms a moving image file, and also supports vehicle information such as accelerator opening and traveling speed to the moving image. The operation data is converted into data and the recorded data is stored in the storage unit 104. In addition, the recording unit 103 may acquire navigation information as necessary and make it correspond to data as driving data. Note that the moving image may be a moving image such as MPEG, or may be composed of a large number of still images such as JPEG.
The storage unit 104 is composed of a semiconductor memory device and is portable. A moving image or the like is input to the driving evaluation system 1 using the storage unit 104. A portable HDD device or a rewritable optical disk may be used. Further, a moving image or the like may be input to the driving evaluation system 1 via a communication interface such as a wireless LAN interface or a USB interface.

  In this way, data such as a moving image is recorded by an in-vehicle device mounted on the vehicle. The moving image is preferably recorded so as to be based on the viewpoint of the driver during driving. Note that a moving image is not limited to only one imaging range. A plurality of ranges during driving (for example, front, rear, in-vehicle, etc.) may be recorded and handled integrally.

  In addition, in each collected moving image, a reference object accompanied by a traffic rule such as a sign that the driver should recognize during driving is recorded as a video.

  Next, a processing operation example of the driving evaluation system 1 will be described.

Moving image data during driving of individual drivers obtained from many on-vehicle devices 100 is sequentially stored in the moving image storage unit 10.
The driving evaluation system 1 starts image recognition processing for the stored moving image data automatically or in response to an operation by a human.

In response to the start of the image recognition process, the image recognition processing unit 30 extracts a reference object used for driving evaluation from the received moving image data, and generates an evaluation log using the extracted position based on the extracted reference object as log contents. Process and output to the log storage unit 20.
This evaluation log reflects traffic regulations such as road signs and temporary stops. As a result of this processing, for example, a log file recording what kind of road sign or temporary stop line can be identified at which time is obtained. In other words, it is recorded at which time the reference object that generates the traffic rule used for evaluation can be identified. Further, the point at which the reference object is identified may be recorded in the log.
A detailed image recognition processing example performed by the image recognition processing unit 30 will be described later.

  The log visualization unit 40 uses the moving image data and the corresponding evaluation log to the evaluator to process the evaluation screen that visualizes the log shorter than the time recorded in the moving image data. Present.

  The evaluator who confirms the screen for evaluation refers to the video displayed on the screen and the traffic rules generated by the reference object used for driving evaluation and the current situation (the visibility and brightness of the reference object, the surroundings) The vehicle driving (vehicle speed, brake timing, etc.) is evaluated appropriately.

  This evaluation time is shortened in time compared with the case where the log visualization unit 40 confirms the moving image as it is. This shortening is a time-based shortening process in which a shortened section (fast forward section, skip section) is defined with an emphasis on the timing at which a reference object appears, based on an evaluation time that can be applied to one moving image data (one driver). It may be based on, or may be shortened as appropriate according to the extraction location (number of extractions) with emphasis on the timing at which the reference object appears. Further, the reference object to be used (evaluated) may be arbitrarily selected. It is desirable that this selection can be made according to the position and type of the reference object.

  In consideration of the evaluator's proficiency level and the like, it is possible to shorten the time required for fast-forwarding, frame-by-frame advancement, etc. as necessary in a section where a reference object to be emphasized appears as well as a shortened section where the reference object does not appear.

Thus, according to the present embodiment, the efficiency of the visual confirmation work can be improved.
For this reason, the number of processing of moving images per time increases, and the accuracy of confirmation by an evaluator such as a sign that each driver included in each collected moving image should recognize during driving is improved.

  In addition, drive using the scope (legal rules and private rules) that should be evaluated, including the fact that the signs are not visible (shaded or damaged by other vehicles, etc.) It can be reflected in the evaluation.

  Here, a processing example of image recognition performed by the image recognition processing unit 30 and a screen display example output from the log visualization unit 40 will be described.

The image recognition process shown in FIG. 3 is a process for identifying a pause point by the image recognition engine.
As shown in the figure, the processing unit by the image recognition engine sequentially acquires images one by one from the input moving image, and executes the “stop” sign detection process and the stop line detection process in parallel, The position to be stopped is identified by integrating both detection results.

  As a processing result, a flag that can identify the reference object and the extraction position is recorded if the temporary stop point can be identified. This flag may be associated with a position in the image or a range (time or distance) covered by traffic rules, or information may be given by other methods.

  What is necessary is just to perform in parallel along the content which reflects such an image identification process on evaluation. What is necessary is just to select the process content to extract so that the reference | standard object relevant to the evaluation item regarded as important may be extracted. Further, the processing content may be determined by arbitrarily selecting a position (range) on a reference target image that is extracted from a moving image by a human. For example, the amount of processing resources can be adjusted by identifying a pedestrian as a reference object in the vicinity of a pedestrian crossing or in the vicinity of a park entrance and not performing recognition elsewhere.

  In addition, it is desirable to perform identification processing of a plurality of temporary stop points in parallel as the identification processing of the temporary stop points to improve the identification accuracy. For example, together with the above identification method, the road surface sign and the stop line can be identified to increase the identification accuracy of the temporary stop point. Further, at the time of image identification, it may be processed so that an accident, an emergency vehicle, or the like can be selected as a reference object using sound recorded simultaneously with the moving image.

FIG. 4 is an explanatory diagram illustrating a screen display example output from the log visualization unit 40.
As shown in the drawing, on the evaluation screen, a reference object (a stop line in the main image) and the like are overlaid on the original image and highlighted by image recognition by the log visualization unit 40.
This reference object corresponds to the evaluation log obtained by image identification, and flag information that is determined whether to stop or not based on the detection result of the “stop” sign and “stop line” It is tied with.

  This display screen is a screen in which the stop line is at a predetermined position (predetermined distance), and flag information (operation / operation to be performed: priority evaluation item) to be illustrated is displayed based on the flag.

  Moreover, in embodiment mentioned later, the driving | operation operation actually performed by the log visualization part 40 is also shown.

  In addition, as an operation input to the evaluation screen, image frame advance / return function, fast forward (automatic fast forward to the next flag) function, specified timing (set by ini file etc.) so that the evaluator can easily see the processing result May be provided as appropriate. At this time, if there is a designated evaluation time, the log visualization unit 40 performs a re-time reference shortening process for adjusting the evaluation time of the unreproduced section so that the time required for stopping or rewinding is recovered. It is good to do. In other words, the reproduction processing time is readjusted with an emphasis on the timing at which the remaining reference objects appear so as to end at the designated evaluation time. At this time, prioritization of reference objects may be registered in advance, and a process of adjusting a reference object to be retained and a reproduction speed based on the priority may be performed.

  By operating the driving evaluation system 1 in this manner, the driving evaluation time can be accurately shortened while maintaining the evaluation accuracy. In other words, many driving evaluations can be performed while suppressing various costs.

Next, a second embodiment will be described.
In the second embodiment, a driving evaluation system configured to collect moving images obtained from in-vehicle devices mounted on a large number of vehicles and perform visual safety evaluation more easily for all of them will be described.

  FIG. 5 is a block diagram showing a configuration of the second embodiment. As illustrated, in the driving evaluation system 2 of the second embodiment, a driving data storage unit 50, an evaluation input unit 60, and an evaluation storage unit 70 are added to the driving evaluation system 1.

  The driving data storage unit 50 stores driving data corresponding to the moving image stored in the moving image storage unit 10 while the moving image of the vehicle obtained from the in-vehicle device or the like is recorded.

  The log visualization unit 40 obtains driving data obtained from the vehicle while the moving image of the vehicle from which the moving image was acquired is recorded, and corresponds to a screen that is output for evaluation as it is or after processing the content. To present. In other words, the display of operations (acceleration and deceleration, braking, horn, sudden handle, etc.) performed during driving generated based on the corresponding driving data is presented together with the moving image.

  The evaluation input unit 60 receives an evaluation input from the evaluator in association with the moving image presented at the time of evaluation, and stores it in the evaluation storage unit 70.

  The evaluation storage unit 70 stores the evaluation result for each driver performed by the evaluator, and uses it for printing on a form, displaying the evaluation result screen, and the like as necessary.

  By configuring the driving evaluation system 2 in this way, it is possible to determine the position of each time, vehicle speed, accelerator / brake state, weather, temperature, etc. from the functions of the drive recorder or other sensors, in-vehicle computers, etc. Using the driving data obtained from the car, the evaluator can directly recognize the operation performed during driving and use it for the evaluation.

Next, a third embodiment will be described.
In the third embodiment, moving images obtained from in-vehicle devices mounted on a large number of vehicles are collected and configured so that visual safety evaluation can be performed more easily for all of them. A driving evaluation system that performs evaluation and complements the difference in evaluation results for each evaluator will be described.

  FIG. 6 is a block diagram illustrating a configuration of the third embodiment. As shown in the figure, in the driving evaluation system 3 of the third embodiment, an evaluator evaluation unit 80 and an evaluation correction unit 90 are added to the driving evaluation system 2.

  The evaluator evaluation unit 80 reflects each evaluator by reflecting the relationship between the individual evaluation results stored in the log storage unit 20, the operation data storage unit 50, and the evaluation storage unit 70 and the extraction position of the corresponding reference object. Evaluation process. In this evaluator's evaluation process, a method of relatively quantifying the difference in evaluation with respect to the same scene for each evaluator can be used. The evaluation items of the evaluator may be determined as required, such as quantification of the difference in evaluation for similar scenes by the evaluator and extraction of evaluation errors. These evaluations use the above-mentioned information to convert the evaluation result part of the part made based on the reference object recognized by the image recognition processing unit 30 into an exemplary evaluation or an evaluation difference with respect to a similar event. It is good also as automatic evaluation which judges based on. Moreover, the person who evaluates the evaluator with reference to the evaluation made by the evaluator can also perform the evaluation.

  The evaluation correction unit 90 corrects the evaluation results of individual driving based on the evaluation of the evaluator performed by the evaluator evaluation unit 80 and records the correction in the evaluation storage unit 70. By this processing, the subjective difference caused by different evaluators can be corrected. The evaluator's evaluation is based on an evaluation that reflects the relationship between the same driving data and the extraction position of the reference object by allowing the evaluator to evaluate the reference case in advance. You may ask for.

  By configuring the driving evaluation system 3 in this way, the system operator can collect and operate the evaluation results in which the difference between the evaluation results is reduced.

  Note that the image recognition processing unit 30 may add a flag indicating an important point indicating that this image or a predetermined amount before the image is reproduced to the image position where the image can be identified.

  Moreover, you may add the flag which shows the kind, the number, and the position of the sign and stop line which could be identified, respectively. Further, it may be displayed on a map in conjunction with navigation information, or may be linked with data from other vehicles.

  Next, the present invention will be described by showing a plurality of examples.

[Example 1]
In this driving evaluation system, the driver of each business vehicle passes a moving image taken during driving of the vehicle at the end of the job to the manager side and receives driving evaluation on a daily basis.
Further, the administrator side assigns an evaluator and efficiently performs the driving evaluation for each individual based on the large number of transferred moving images.

As shown in FIG. 7, the driving evaluation system is divided into a host server and a workstation having an image recognition processing unit 30, a log visualization unit 40, etc., and after driving evaluation for each driver, Provide communication means to send to the server. In this embodiment, the driving evaluation system 3 is configured.
In this system, as a configuration management, an evaluation item that is regarded as important is notified from a higher-level server to a workstation at each branch office, and is reflected in the reference object (image identification process to be used) and the evaluation item to be extracted. Each reference object is set. In the configuration management, hardware and software configurations and version management are performed for workstations at branch offices and individual drive recorders. It may also manage whether the workstation load is appropriate. Further, for example, parameter optimization may be performed in which processing of an arbitrary rule is assigned for each region.

At the branch office workstation, the moving image storage unit 10 sequentially receives a large number of various types of information acquired by individual vehicles.
The image recognition processing unit 30 extracts a reference object or the like used for driving evaluation from the recorded moving image, generates a log, and stores it in the log storage unit 20.
The evaluator inputs individual evaluation results to the evaluation input unit 60 with reference to the evaluation screen presented from the log visualization unit 40.

The daily evaluation result stored in the evaluation storage unit 70 can be confirmed by printing it on a form or the like.
The image data is stored and searched for each driver and each vehicle.

  All of these processes are evaluated by a department that manages business vehicles such as branch offices. Thereafter, only the evaluated operation result is sent to the host server. The host server collects the analysis and evaluation results of individual drivers, performs overall operation and analysis management for each branch office, and uses it to reduce future accidents. Further, the difference between the evaluator's evaluation and the evaluation for each evaluator may be corrected on the headquarters side.

  By separating the upper server and the lower device in this way, it is possible to distribute the image recognition process as appropriate and efficiently perform the driving evaluation for each individual with respect to a large number of vehicles.

[Example 2]
In the driving evaluation system according to the second embodiment, as in the first embodiment, a large number of passed moving images are efficiently driven by the evaluator's visual observation using moving images taken during driving of the vehicle. Evaluate.
The difference from the first embodiment is that the image recognition processing unit 30 and the log visualization unit 40 are provided on the higher server side.

In the second embodiment, as shown in FIG. 8, the driving evaluation system is constructed by dividing it into an upper server having a log visualization unit 40 and a workstation as a lower device. For example, in a department that manages business vehicles such as a branch office, related data to be evaluated such as recording data and driving data is sent to a host server. The host server receives relevant data to be evaluated, performs driving evaluation of each driver, and sends the evaluation result for each driver to a workstation at each branch office.
The host server performs evaluation work, collects individual analysis evaluation results, and performs overall analysis management.

  Thus, by separating the upper server and the lower apparatus, the image recognition process can be performed in a centralized manner.

[Example 3]
In the driving evaluation system according to the third embodiment, as in the first and second embodiments, a large number of moving images passed by the evaluator can be visually observed using moving images taken during driving of the vehicle. Evaluate driving efficiently.
The difference from the second embodiment is that an image recognition processing unit 30 is provided on the workstation side.

  In the third embodiment, as shown in FIG. 9, the driving evaluation system is constructed by dividing it into a host server having a log visualization unit 40 and a workstation having an image recognition processing unit 30. For example, in a department that manages business vehicles such as a branch office, image recognition processing is performed, and related data to be evaluated, such as recorded data, an evaluation log, and driving data, is sent to a host server. The host server receives relevant data to be evaluated, performs driving evaluation of each driver, and sends the evaluation result for each driver to a workstation at each branch office. The host server performs evaluation work by the evaluator, collects individual analysis evaluation results, and performs overall analysis management.

  Thus, by separating the upper server and the lower device, distributed processing of image recognition processing can be performed.

[Example 4]
In the driving evaluation system according to the fourth embodiment, as in the first to third embodiments, a large number of moving images passed by an evaluator can be visually observed using moving images taken during driving of the vehicle. Evaluate driving efficiently.
The difference from the other embodiments is that the workstation has an image recognition processing unit 30 and a log visualization unit 40, and analysis management is also performed on the workstation side.

  In the fourth embodiment, as shown in FIG. 10, the driving evaluation system is constructed by dividing it into a workstation having an image recognition processing unit 30, a log visualization unit 40, and the like and a host server that performs configuration management. Each process is individually performed at a workstation at each branch office, and the host server is configured to collectively perform only configuration management such as update and maintenance management of each branch office via the management terminal.

  By separating the upper server and the lower device in this way, it is possible to carry out from image recognition processing to evaluation by the evaluator according to individual circumstances for each branch office.

  In addition, what is necessary is just to implement | achieve each part of a driving | operation evaluation system using the combination of hardware and software. In a form in which hardware and software are combined, an operation evaluation program is expanded in the RAM, and each unit is realized as various means by operating hardware such as a control unit (CPU) based on the program. Further, the driving evaluation program may be recorded on a storage medium and distributed. The program recorded on the recording medium is read into a memory via a wired, wireless, or recording medium itself, and operates a control unit or the like. Examples of the recording medium include an optical disk, a magnetic disk, a semiconductor memory device, and a hard disk.

  To describe the above embodiment in another expression, an information processing apparatus that operates as a driving evaluation system is based on a driving evaluation program developed in a RAM, image recognition means, log storage means, moving image storage means, log visualization It can be realized by operating the control unit as means.

  Although the embodiments and examples have been illustrated and described above, changes such as separation / merging of block configurations and replacement of procedures are free as long as the gist of the present invention and the functions to be described are satisfied. The invention is not limited.

  As described above, the driving evaluation system to which the present invention is applied can provide an environment in which driving evaluation performed with reference to a moving image can be performed at low cost.

  Further, in the present invention, by obtaining and providing traffic regulations that are not fully computerized by image recognition, it is possible to compensate for evaluator evaluation omissions, evaluation blurring, subjective evaluation differences among evaluators, etc. Can do.

  In addition, the specific configuration of the present invention is not limited to the above-described embodiment, and changes within a range not departing from the gist of the present invention are included in the present invention.

  The present invention is suitable for a transportation business that manages a large number of business vehicles and a system in a taxi business.

1, 2, 3 Driving evaluation system 10 Moving image storage unit (moving image storage means)
20 Log storage unit (log storage means)
30 Image recognition processing unit (image recognition processing means)
40 Log visualization part (log visualization means)
50 Operation data storage (operation data storage means)
60 Evaluation input section (Evaluation input means)
70 Evaluation storage unit (evaluation storage means)
80 Evaluator Evaluation Department (Evaluator Evaluation Method)
90 Evaluation Correction Unit (Evaluation Correction Unit)

Claims (20)

Moving image storage means for storing a moving image recorded by an imaging means mounted on the vehicle;
An image recognition unit that extracts a reference object used for driving evaluation from a moving image recorded in the moving image storage unit and outputs at least an extraction position;
Log storage means for storing the extraction position output by the image recognition means as an evaluation log used for driving evaluation;
When evaluating the moving image recorded in the moving image storage means, based on at least the extraction position of the reference object stored in the log storage means, while leaving a section in which the reference object included in the moving image appears , by shortening the regeneration time interval, the log visualization means for reproducing processes applicable moving image by reducing the reproduction processing time of the entire moving image to be evaluated,
With
A driving evaluation system characterized in that driving evaluation of a recorded moving image can be performed visually. The log visualization means obtains driving data obtained from the vehicle side while the moving image of the vehicle from which the moving image was acquired is recorded, and when performing visual evaluation, corresponding driving data together with the reproduced moving image The driving evaluation system according to claim 1, wherein an operation performed during driving generated based on the above is presented together with a moving image. The log visualization means presents, together with a moving image, a highlight display indicating a reference object and / or an evaluation reference position extracted from the moving image when visual evaluation is performed. The driving evaluation system described in 1. The system comprises a lower device having at least the log visualization means, and an upper server,
The lower device comprises a communication means for sending an evaluation result to the upper server after performing a driving evaluation for each driver using the log visualization means ,
The driving evaluation system according to any one of claims 1 to 3, wherein the upper server collects evaluation results of individual drivers from the lower device and performs overall analysis management. In the system, comprising a host server having at least the log visualization means,
The upper server receives relevant data to be evaluated, and can perform an individual driver's driving evaluation visually from the corresponding moving image using the log visualization means, and collects the evaluation results of the individual drivers. The operation evaluation system according to any one of claims 1 to 3, wherein analysis management as a whole is performed. In the system, comprising: a lower apparatus having at least the image recognition means and the log storage means; and an upper server having at least the log visualization means ,
The sub system is provided with a communication unit for sending the log stored in said log storage means to obtain an image recognition processing by the image recognition means to said host server,
The upper server receives relevant data to be evaluated, and can perform an individual driver's driving evaluation visually from the corresponding moving image using the log visualization means, and collects the evaluation results of the individual drivers. The operation evaluation system according to any one of claims 1 to 3, wherein analysis management as a whole is performed. The log visualization means, based on at least the extraction position of the reference object and the priority level of the reference object stored in the log storage means during the reproduction process when evaluating the moving image in response to the designation of the evaluation time Thus, the playback processing time of the entire moving image to be evaluated is reduced by shortening the playback time of the other sections while preferentially leaving the section in which the reference object having a higher priority level included in the moving images of the unplayed section appears. The driving evaluation system according to any one of claims 1 to 6, wherein an adjustment is performed so as to approximate the specified evaluation time.   An evaluation by an evaluator who visually evaluates the driving performance of each individual driver by automatically reflecting each evaluation result and the relationship between the driving data used in the evaluation result and the extraction position of the reference object The driving evaluation system according to claim 1, further comprising a person evaluation unit.   9. An evaluation correction unit that corrects the driving evaluation of each driver based on the evaluation of the evaluator reflecting the relationship between the driving data and the extraction position of the reference object. The driving | operation evaluation system as described in any one of these. Storing a moving image recorded by an imaging means mounted on the vehicle;
Extracting a reference object used for driving evaluation from the recorded moving image by image recognition processing, and storing at least an extraction position as an evaluation log used for driving evaluation;
When evaluating the recorded moving image, based on at least the extraction position of the reference object recorded in the evaluation log, the reproduction time of other sections is set while leaving the section in which the reference object included in the moving image appears. A process of shortening the reproduction processing time of the entire moving image to be evaluated by shortening and reproducing the corresponding moving image;
A driving evaluation method characterized by causing the information processing system to perform a driving evaluation of a recorded moving image. Further comprising a step of storing driving data obtained from the vehicle side while the moving image of the vehicle that acquired the moving image is recorded;
The driving evaluation method according to claim 10, wherein, when visual evaluation is performed, an operation performed during driving generated based on corresponding driving data together with a moving image to be reproduced is presented together with the moving image. . The driving evaluation method according to claim 10 or 11, wherein when visual evaluation is performed, highlighting indicating a reference object and / or an evaluation reference position extracted from a moving image is presented together with the moving image. .   After the visual evaluation, the evaluation by the evaluator who performed the driving evaluation of each driver visually was performed.The relationship between the individual evaluation results, the driving data used in the evaluation results, and the extraction position of the reference object was determined. The driving evaluation method according to any one of claims 10 to 12, further comprising a step of performing automatic determination with reflection.   The method further comprises a step of correcting the driving evaluation of each driver based on the evaluation of the evaluator reflecting the relationship between the driving data and the extraction position of the reference object after performing the visual evaluation. The driving | running evaluation method as described in any one of Claim 10 thru | or 13. Control the information processing system
Moving image storage means for storing a moving image recorded by an imaging means mounted on the vehicle;
An image recognition unit that extracts a reference object used for driving evaluation from a moving image recorded in the moving image storage unit and outputs at least an extraction position;
Log storage means for storing the extraction position output by the image recognition means as an evaluation log used for driving evaluation;
When evaluating the moving image recorded in the moving image storage means, based on at least the extraction position of the reference object stored in the log storage means, while leaving a section in which the reference object included in the moving image appears , by shortening the regeneration time interval, the log visualization means for reproducing processes applicable moving image by reducing the reproduction processing time of the entire moving image to be evaluated,
As an operation evaluation program, the operation evaluation is performed so that the operation evaluation of the recorded moving image can be performed visually. When the log visualization means obtains driving data obtained from the vehicle side while the moving image of the vehicle from which the moving image is acquired is recorded and performs visual evaluation, the corresponding driving data is reproduced together with the reproduced moving image. 16. The driving evaluation program according to claim 15, wherein the driving evaluation program is operated so as to present together with the moving image the operation performed based on the driving. The log visualization means is operated so as to present, together with a moving image, an emphasis display indicating a reference object and / or an evaluation reference position extracted from a moving image when visual evaluation is performed. Item 15. The driving evaluation program according to Item 15 or 16. Control the information processing system
An evaluation by an evaluator who visually evaluates the driving performance of each individual driver by automatically reflecting each evaluation result and the relationship between the driving data used in the evaluation result and the extraction position of the reference object The driving evaluation program according to any one of claims 15 to 17, wherein the driving evaluation program is further operated as a person evaluation means. Control the information processing system
The driving evaluation of each driver is further operated as an evaluation correcting means for correcting based on the evaluation of the evaluator reflecting the relationship between the driving data and the extraction position of the reference object. The driving | running evaluation program as described in any one of 18. When the log visualization means evaluates the moving image in response to the designation of the evaluation time, it is based on at least the extraction position of the reference object stored in the log storage means and the priority level of the reference object in the reproduction process. Thus, the playback processing time of the entire moving image to be evaluated is reduced by shortening the playback time of the other sections while preferentially leaving the section in which the reference object having a higher priority level included in the moving images of the unplayed section appears. The driving evaluation program according to any one of claims 15 to 19, wherein the driving evaluation program is caused to function so as to perform an adjustment to approach the designated evaluation time.

Images