JP6721022B2 - Image compression device and program - Google Patents

Description

The present invention relates to an image compression device and a program.

A technique of generating information for quantitatively evaluating block distortion of a compressed image based on an original image and a reproduced image obtained by decoding a compressed image obtained by compressing the original image is disclosed (for example, see Patent Document 1). ).

JP 05-219384 A

The image quality of the compressed image changes under the influence of the external environment such as the amount of movement of the imaging target and the brightness of the imaging region including the imaging target. The optimum value of the compression parameter that defines the condition of image compression varies depending on the amount of movement of the imaging target and the brightness of the imaging region. Therefore, when the imaging target is imaged in a place affected by the external environment, there is a problem that the image quality of the compressed image may deteriorate due to the change in the external environment.

Therefore, the present invention has been made in view of these points, and an object of the present invention is to provide an image compression apparatus that compresses an image so that the image quality of the image after compression becomes a desired image quality.

An image compression apparatus according to a first aspect of the present invention is an acquisition unit that acquires image data, a compression unit that compresses the image data based on a predetermined compression parameter, and a compressed image after the compression unit has compressed the image data. An image quality evaluation unit that evaluates image quality of data, a data amount of the compressed image data, and a similarity between image data before compression by the compression unit and compressed image data after compression by the compression unit. And a parameter adjusting unit that adjusts the compression parameter used by the compression unit by specifying a characteristic of the compressed image data based on the above and selecting a compression parameter associated with the specified characteristic. ..

When the data amount of the compressed image data is equal to or more than a first data threshold and the value of the mean square error indicating the similarity is equal to or more than a first evaluation threshold, the parameter adjusting unit is configured to The feature of the compressed image data may be specified to be rainy weather.

The parameter adjusting unit specifies that the features of the image data and the compressed image data are fine weather when the data amount of the compressed image data is equal to or less than a second data threshold that is smaller than the first data threshold. May be.

When the data amount of the compressed image data is equal to or larger than a first data threshold and the value of the mean square error is equal to or smaller than a second evaluation threshold smaller than the first evaluation threshold, The features of the image data and the compressed image data may not be specified.

The parameter adjustment unit may specify the characteristics of the compressed image data based on the tendency of the data amount of the compressed image data to increase or decrease. The parameter adjustment unit may specify the characteristics of the compressed image data based on the tendency of the similarity to increase or decrease.

The program according to the second aspect of the present invention is a program that causes a computer to acquire an image data, a compression unit that compresses the image data based on a predetermined compression parameter, and compressed image data after the compression unit has compressed the image data. An image quality evaluation unit for evaluating the image quality of, and a data amount of the compressed image data, and a similarity between the image data before being compressed by the compressing unit and the compressed image data after being compressed by the compressing unit. , And specifies a feature of the compressed image data, and selects a compression parameter associated with the identified feature to function as a parameter adjustment unit that adjusts the compression parameter used by the compression unit.

According to the present invention, there is an effect that an image can be compressed so that the image quality of the compressed image becomes a desired image quality.

It is a figure for explaining the outline of a video recording system. It is a figure which shows the function structure of an image compression apparatus. It is a figure which shows typically the graph which plotted the evaluation result which evaluated the image after compression for every frame. It is a figure for demonstrating the process which acquires target information linked|related with some area|regions. It is a figure for demonstrating the change of an evaluation result. It is a flowchart of the process which adjusts a compression parameter. It is a figure which shows the function structure of the image compression apparatus which concerns on 2nd Embodiment. It is a figure which shows typically the data table which shows the accumulated evaluation result. It is a figure which shows typically the data table which linked|related the evaluation result, the characteristic of image data, and the used compression parameter.

<First Embodiment>
FIG. 1 is a diagram for explaining the outline of the video recording system S. The video recording system S includes a camera 2 that captures an imaging target, an image compression device 1 that compresses a captured image generated by the camera 2, and a video storage device 3. The image compression device 1 and the camera 2 are connected via a network. Similarly, the image compression device 1 and the video storage device 3 are connected via a network. The network is, for example, the Internet.

The image compression apparatus 1 acquires an image obtained by capturing an object to be captured by the camera 2 and generates a post-compression image compressed by a predetermined compression parameter. The image compression device 1 transmits the generated compressed image to the video storage device 3. The video storage device 3 stores the compressed image received from the image compression device 1 in a storage medium such as a hard disk. A person who monitors the area imaged by the camera 2 (hereinafter referred to as user) can confirm the state of the area imaged by the camera 2 by visually recognizing the compressed image stored in the video storage device 3.

The image quality of the compressed image compressed by the image compression apparatus 1 changes depending on the external environment such as the amount of movement of the image capturing target when the camera 2 captures the image capturing target and the brightness in the area including the image capturing target. Therefore, in order to obtain the desired image quality of the image after compression, it is necessary to adjust the compression parameter that defines the condition of image compression according to the imaging target and the external environment. The desired image quality is, for example, a target of image quality set by a user who monitors the situation of an area including an imaging target by using the compressed image.

Therefore, the image compression device 1 adjusts the compression parameter based on the difference between the evaluation result of evaluating the image quality of the compressed image after compression and the predetermined image quality target. Specifically, the image compression device 1 adjusts the compression parameter so that the difference between the evaluation result and the target becomes small. By doing so, the image compression apparatus 1 can make the image quality of the compressed image a desired image quality by adjusting the compression parameter even when the imaging target and the external environment change.

FIG. 2 is a diagram showing a functional configuration of the image compression device 1. The image compression device 1 includes an acquisition unit 11, a compression unit 12, an image quality evaluation unit 13, a target information acquisition unit 14, and a parameter adjustment unit 15. The acquisition unit 11 acquires the image data V0 from the camera 2. The acquisition unit 11 notifies the compression unit 12 and the image quality evaluation unit 13 of the acquired image data V0.

The compression unit 12 compresses the image data V0 based on a predetermined compression parameter to generate the compressed image data V1. The compression standard used when the compression unit 12 compresses image data is, for example, H.264. H.264, but is not limited to this. The compression parameters are various setting values that determine the image quality of the image after compression. Specifically, the compression parameter is H.264. In the case of H.264, it is the number of macroblocks per second, the number of macroblocks per frame, and the bit rate. Further, the compression parameter may include a GOP (Group Of Picture) and an I frame (Key frame) insertion interval.

The image quality evaluation unit 13 evaluates the image quality of the compressed image data V1 after being compressed by the compression unit 12. For example, the image quality evaluation unit 13 calculates a mean square error (Mean Square Error: MSE) between the uncompressed image data V0 and the compressed image data V1 after the compression unit 12 compresses the compressed image data V0. The image quality of the data V1 is evaluated. The image quality evaluation unit 13 evaluates that the lower the MSE value, the higher the image quality.

The image quality evaluation unit 13 may evaluate the image quality level based on the MSE value. For example, the image quality evaluation unit 13 evaluates the image quality level in 10 levels. The image quality evaluation unit 13 evaluates, for example, MSE of less than 50 as level 0, MSE of 50 or more and less than 100 as level 1, and MSE of 100 or more and less than 150 as level 2. The image quality evaluation unit 13 increases the level by one level each time the MSE increases by 50, and evaluates the MSE of 400 or more and less than 450 as the level 8 and the MSE of 450 or more as the level 9.

FIG. 3 is a diagram schematically showing a graph in which evaluation results obtained by evaluating the compressed image for each frame are plotted. In FIG. 3, the horizontal axis represents time and the vertical axis represents the evaluation result (value of MSE). In FIG. 3, the target image quality level of the evaluation result determined by the target indicated by the target information is indicated by a broken line. Further, in FIG. 3, the time indicated by arrows I0 and I1 is the timing of the I frame. Since the I frame is not affected by the data error in the other frames, the evaluation result is better than that of the other frames.

In FIG. 3, the evaluation result R1 and the evaluation result R2 indicated by solid lines are the results of evaluating the compressed image obtained by compressing the image captured by the camera 2 for each frame. The evaluation result R1 is an evaluation result obtained by evaluating a compressed image obtained by compressing a video with little change between images. The image quality evaluation unit 13 evaluates the evaluation result R1 as the image quality level 1.

The evaluation result R2 is an evaluation result obtained by evaluating a compressed image obtained by compressing a video having a large change between images. The image quality evaluation unit 13 evaluates the evaluation result R2 as the image quality level 3 based on the average MSE value obtained by averaging the MSE values within the predetermined average period, which is indicated by the alternate long and short dash line M in FIG. 3, for example. The average period is, for example, a period from the I frame to the next I frame, but is not limited to this, and may be appropriately set by a business operator who operates the video recording system S.

The image quality evaluation unit 13 may evaluate the image quality level based on the maximum value of MSE values within the average period. When evaluating the image quality level based on the maximum value of the MSE values within the average period, the image quality evaluation unit 13 evaluates the evaluation result M2 as the image quality level 4.

The image quality evaluation unit 13 may evaluate the image quality not only by calculating the MSE but also by calculating a peak signal-to-noise ratio (Peak signal-to-noise ratio) or an SSIM (structural similarity). The image quality evaluation unit 13 notifies the parameter adjustment unit 15 of the calculated evaluation result.

The target information acquisition unit 14 acquires target information indicating a target of image quality of the compressed image data V1. The target information is, for example, text or a numerical value indicating an image quality level indicating a target of image quality input by the user. In the following description, it is assumed that the target information acquisition unit 14 has acquired 2 as the numerical value indicating the image quality level input by the user. The target information acquisition unit 14 may determine the target value of the MSE value based on the numerical value indicating the image quality level input by the user.

The target information acquisition unit 14 may acquire the target information in association with at least a part of the plurality of regions of the image data V0. FIG. 4 is a diagram for explaining a process of acquiring target information in association with a part of the area. In FIG. 4, it is assumed that the image data A is divided into six areas A1 to A6. The user can select at least one of the six areas A1 to A6. The target information acquisition unit 14 acquires information indicating the area selected by the user.

The target information acquisition unit 14 notifies the parameter adjustment unit 15 of the acquired target information. Further, when the target information acquisition unit acquires the target information in association with a part of the area, the target information acquisition unit notifies the parameter adjustment unit 15 of the acquired target information and the information indicating the partial area selected by the user.

The parameter adjustment unit 15 adjusts the compression parameter based on the difference between the evaluation result of the image quality evaluation unit 13 and the target indicated by the target information. The parameter adjustment unit 15 changes the compression parameter so that the difference between the evaluation result and the target becomes small, for example. Specifically, the parameter adjusting unit 15 adjusts the compression parameter so that the difference between the image quality level corresponding to the evaluation result and the image quality level indicated by the target becomes small. For example, the parameter adjusting unit 15 adjusts the value of the compression parameter so that the value obtained by subtracting the target value of MSE determined by the target from the value of MSE that is the evaluation result becomes smaller. Further, the parameter adjusting unit 15 may not adjust the compression parameter as long as the value of MSE is equal to or less than the target value.

FIG. 5 is a diagram for explaining changes in the evaluation result. In FIG. 5, the horizontal axis represents time, the vertical axis represents the evaluation result (value of MSE), and the image quality level determined by the target indicated by the target information is shown by a broken line. The evaluation result M1 indicated by the solid line is a graph in which the average value of MSE values is plotted against time.

The parameter adjusting unit 15 adjusts the compression parameter at time t _i so that the difference between the evaluation result (image quality level 7) and the target (image quality level 2) in the period T1 becomes small. Similarly, the parameter adjusting unit 15 adjusts the compression parameter at time t _i+1 so that the difference between the evaluation result (image quality level 4) in the period T2 and the target (image quality level 2) becomes small. The parameter adjustment unit 15 adjusts the compression parameter at time t _i+2 so that the difference between the evaluation result (image quality level 3) and the target (image quality level 2) in the period T3 becomes small. At time t _i+3 , the parameter adjusting unit 15 does not adjust the compression parameter because there is no difference between the evaluation result (image quality level 2) in the period T4 and the target (image quality level 2). By doing so, the parameter adjusting unit 15 can compress the image with the image quality desired by the user.

In addition, the parameter adjusting unit 15 may adjust the compression parameter based on the difference between the evaluation result of the image quality of the compressed image data V1 of at least a part of the area where the target information acquiring unit 14 has acquired the target information and the target. Good. The parameter adjusting unit 15 adjusts the compression parameter so that the difference between the evaluation result of at least a part of the area and the target associated with the part of the area is small. Specifically, the parameter adjusting unit 15 adjusts the compression parameter so that the value of MSE that is the evaluation result of a part of the area is included in the target range of the MSE that is determined by the target associated with the part of the area. To do.

When the target information acquisition unit 14 acquires, for example, the target associated with the area A5, the parameter adjustment unit 15 includes the result of the image quality evaluation performed by the image quality evaluation unit 13 and the target associated with the area A5. Adjust the compression parameters so that the difference is small. In addition, when the target information acquisition unit 14 acquires the target associated with the area A5 and the target associated with the area A6, the parameter adjustment unit 15 acquires the difference between the evaluation result of the area A5 and the target, and the area A6. The compression parameter is adjusted so that the difference between the evaluation result and the target becomes small. The parameter adjusting unit 15 adjusts the compression parameter so that the MSE value of the area A5 is included in the target range of the area A5 and the MSE value of the area A6 is included in the target range of the area A6, for example.

The parameter adjusting unit 15 may adjust the compression parameter so that the difference between the average of the evaluation results of the image quality evaluated in each of the plurality of areas and the target becomes small. Specifically, the parameter adjusting unit 15 sets a predetermined difference between the average value of the MSE value that is the evaluation result of the area A5 and the MSE value that is the evaluation result of the area A6, and the target value of the MSE that is determined by the target. The compression parameter is adjusted so as to be equal to or less than the threshold of. By doing so, the image compressing apparatus 1 can compress the image so that the image quality of the area of interest to the user becomes the image quality desired by the user, and reduce the amount of processing.

FIG. 6 is a flowchart of the process of adjusting the compression parameter. Hereinafter, the flow of processing in which the image compression apparatus 1 adjusts the compression parameter will be described with reference to FIG. First, the acquisition unit 11 acquires the image data V0 obtained by capturing the image of the image capturing target by the camera 2 (step S1). The compression unit 12 compresses the image data V0 acquired by the acquisition unit 11 to generate post-compression image data V1 (step S2). The image quality evaluation unit 13 evaluates the image quality of the compressed image data V1 (step S3).

Subsequently, the target information acquisition unit 14 acquires target information indicating the image quality target of the compressed image data V1 input by the user (step S4). Then, the parameter adjusting unit 15 adjusts the compression parameter based on the difference between the evaluation result and the image quality target (step S5). The image compression device 1 can adjust the compression parameter of the image data V0 captured by the camera 2 by repeating steps S1 to S5 so that the image quality of the compressed image data V1 becomes the target image quality. .. The image compressing apparatus 1 may execute either the process of step S1 or the process of step S4 first, or the process of steps S1 to S3 and the process of step S4 may be executed in parallel. May be.

[Effects of the image compression device 1 according to the first embodiment]
As described above, the parameter adjustment unit 15 of the video recording system S adjusts the compression parameter based on the difference between the evaluation result of evaluating the image quality of the compressed image data V1 and the target of the image quality of the compressed image data V1. To do. By doing so, the video recording system S can set the image quality of the image compressed by the compression unit 12 to the image quality desired by the user.

<Second Embodiment>
The image compression device 1 may have a configuration for accumulating evaluation results in addition to the above configuration. FIG. 7 is a diagram showing a functional configuration of the image compression apparatus 1 according to the second embodiment. The image compression device 1 according to the second embodiment further includes a storage unit 16 that stores the evaluation result. The storage unit 16 stores the evaluation result in association with the date and time when the image data was captured, for example.

FIG. 8 is a diagram schematically showing a data table showing the accumulated evaluation results. FIG. 8A is a diagram schematically showing a data table showing the evaluation results accumulated in association with the date and time. The graph in which the evaluation results accumulated in association with the date and time are plotted against time is, for example, a graph as shown in FIG.

The parameter adjusting unit 15 adjusts the compression parameter based on the evaluation result accumulated by the accumulating unit 16. The parameter adjusting unit 15 refers to, for example, the evaluation result accumulated in the accumulating unit 16 after the same time as the current time on another day, and compresses so that the difference between the referred evaluation result and the target becomes small. Adjust the parameters. By doing so, the parameter adjusting unit 15 can change the compression parameter in advance in accordance with the tendency that the evaluation result changes.

The storage unit 16 may store the evaluation result in association with the feature of the image data V0 or the compressed image data V1. The feature is, for example, weather such as fine weather, cloudy weather, rainy weather, and snowy weather. The characteristic may be a time zone, brightness, or the like. FIG. 8B is a diagram schematically showing a data table in which image data V0, post-compression image data V1, features of the post-compression image data V1 and evaluation results of the post-compression image data V1 are associated with each other. Is.

When the storage unit 16 stores the feature of the image data V0 or the compressed image data V1 and the evaluation result in association with each other, the parameter adjustment unit 15 is close to the feature of the compressed image data V1 generated by the compression unit 12. Adjust the compression parameters based on the evaluation results associated with the features. The parameter adjustment unit 15 first analyzes the compressed image data V1 generated by the compression unit 12 and specifies the characteristics of the compressed image data V1. The method of identifying the characteristics of the compressed image data V1 is arbitrary, and the details will be described later. The parameter adjustment unit 15 refers to the evaluation result accumulated in association with the specified feature, and thereby, for example, the evaluation result accumulated in the accumulation unit 16 in association with the current time on the other day after the same time. The compression parameter is adjusted so that the difference between the referred evaluation result and the target becomes small.

The storage unit 16 may store the compression parameter used when the evaluation result was obtained. FIG. 9 is a diagram schematically showing a data table in which the evaluation results, the characteristics of the image data V0, and the compression parameters used are associated with each other.

When the storage unit 16 stores the compression parameter in association with the feature of the image data V0, the parameter adjustment unit 15 includes one or more features associated with the same feature as the feature of the compressed image data V1 generated by the compression unit 12. Of the evaluation results, the compression parameter associated with the evaluation result that satisfies the image quality target is selected. For example, when the characteristic of the compressed image data V1 is “sunny”, the parameter adjusting unit 15 associates the evaluation results M11 (image quality level 4) and M12 (image quality level) with the same characteristic “sunny” as the compressed image data V1. In 2), the compression parameter “a” corresponding to the evaluation result M11 that satisfies the image quality target (image quality level 2) is selected. By doing so, the parameter adjusting unit 15 can adjust the compression parameter based on the compression parameter used in the past, so that the time until the image quality of the compressed image becomes the desired image quality is shortened. be able to.

[Method of identifying characteristics of image data]
Hereinafter, a specific example will be given of a method in which the parameter adjusting unit 15 specifies the characteristics of the image data V0 or the compressed image data V1 based on the data amount of the compressed image data V1 and the value of MSE. explain.

Since the image data V0 captured when it is raining contains raindrops, movement occurs in the entire image. Therefore, both the data amount of the compressed image data V1 and the value of MSE become large. Therefore, if the data amount of the compressed image data V1 is equal to or larger than the first data threshold and the value of MES is equal to or larger than the first evaluation threshold, the parameter adjustment unit 15 features the image data V0 and the compressed image data V1. Identify as being rainy.

Since the image data V0 captured when it is sunny does not move much and can be compressed with a high compression rate even with high image quality, both the data amount of the compressed image data V1 and the MES value are small. Therefore, if the data amount of the compressed image data V1 is less than or equal to the second data threshold that is smaller than the first data threshold and the value of MSE is less than the second evaluation threshold that is smaller than the first evaluation threshold, , The image data V0 and the compressed image data V1 are specified to be sunny.

Further, when a sudden motion occurs in the image data V0 captured during sunny weather, the value of MSE increases even though the amount of compressed image data V1 is small. Therefore, if the data amount is equal to or less than the second data threshold and the MES value is equal to or greater than the second evaluation threshold, the parameter adjusting unit 15 identifies that the features of the image data V0 and the compressed image data V1 are fine weather. To do. The parameter adjusting unit 15 does not specify the feature when the data amount is equal to or larger than the first data threshold and the MES value is equal to or smaller than the second evaluation threshold.

The parameter adjustment unit 15 may specify the characteristics of the compressed image data V1 based on the tendency of the data amount of the compressed image data V1 to increase or decrease. For example, when the data amount of the compressed image data V1 and the value of MSE are increasing, the parameter adjusting unit 15 indicates that the weather when the image data V0 was captured changed from fine weather to rainy weather. Specify. Further, when the data amount of the compressed image data V1 and the value of MSE are decreasing, the parameter adjusting unit 15 indicates that the weather when the image data V0 was captured changed from rainy weather to clear weather. Specify.

The parameter adjusting unit 15 selects the compression parameter corresponding to the changed weather, based on the relationship between the compression result and the evaluation result associated with the identified feature corresponding to the changed weather. By doing so, the parameter adjusting unit 15 can quickly adjust the image quality of the compressed image to a compression parameter that makes the image quality desired by the user even when the weather changes.

The parameter adjusting unit 15 may specify the feature by analyzing the image data V0 or the compressed image data V1 using a known analysis method. The analysis method is, for example, a pattern recognition method in which features are specified by selecting and extracting objects having a certain rule or meaning from the image data V0.

[Effects of the image compression device 1 according to the second embodiment]
As described above, the storage unit 16 of the image compression device 1 stores the evaluation result. Then, the parameter adjusting unit 15 adjusts the compression parameter based on the compression parameter corresponding to the accumulated evaluation result. By doing so, the parameter adjusting unit 15 can shorten the time until the image quality of the compressed image reaches the image quality desired by the user.

Although the present invention has been described above using the embodiments, the technical scope of the present invention is not limited to the scope described in the above embodiments, and various modifications and changes are possible within the scope of the gist thereof. is there. For example, the specific embodiment of device distribution/integration is not limited to the above embodiment, and all or part of the device may be functionally or physically distributed/integrated in arbitrary units. You can Further, a new embodiment that occurs due to an arbitrary combination of a plurality of embodiments is also included in the embodiment of the present invention. The effect of the new embodiment produced by the combination also has the effect of the original embodiment.

DESCRIPTION OF SYMBOLS 1 Image compression device 2 Camera 3 Video storage device 11 Acquisition unit 12 Compression unit 13 Image quality evaluation unit 14 Target information acquisition unit 15 Parameter adjustment unit 16 Storage unit

Claims (4)

An acquisition unit that acquires image data,
A compression unit that compresses the image data based on a predetermined compression parameter,
An image quality evaluation unit that evaluates the similarity indicating the value of the mean square error between the image data before compression by the compression unit and the compressed image data after compression by the compression unit,
Based on the data amount of the compressed image data and the degree of similarity evaluated by the image quality evaluation unit, one of a plurality of features indicating the surrounding environment of the image capturing apparatus that captured the image data is specified. Then, a parameter adjusting unit that adjusts the compression parameter used by the compression unit by selecting a compression parameter used in the past, which is associated with the identified feature,
Have a,
When the data amount of the compressed image data is equal to or more than a first data threshold value and the value of the mean square error is equal to or more than a first evaluation threshold value, the parameter adjusting unit may include the image data and the compressed image data. Identify that the feature is rainy weather,
Image compression device. The parameter adjustment unit specifies that the characteristics of the image data and the compressed image data are sunny when the data amount of the compressed image data is equal to or less than a second data threshold that is smaller than the first data threshold. ,
The image compression apparatus according to claim 1 . When the data amount of the compressed image data is equal to or larger than a first data threshold and the value of the mean square error is equal to or smaller than a second evaluation threshold smaller than the first evaluation threshold, Does not specify the characteristics of the image data and the compressed image data,
The image compression apparatus according to claim 1 or 2. Computer,
An acquisition unit that acquires image data,
A compression unit that compresses the image data based on a predetermined compression parameter,
An image quality evaluation unit that evaluates the similarity indicating the value of the mean square error between the image data before compression by the compression unit and the compressed image data after compression by the compression unit, and the data amount of the compressed image data. Based on the similarity evaluated by the image quality evaluation unit, one of a plurality of characteristics indicating the surrounding environment of the image capturing apparatus that has captured the image data is identified and associated with the identified characteristic. A parameter adjusting unit that adjusts the compression parameter used by the compression unit by selecting a compression parameter used in the past ,
To function as,
When the data amount of the compressed image data is equal to or more than a first data threshold value and the value of the mean square error is equal to or more than a first evaluation threshold value, the parameter adjusting unit may include the image data and the compressed image data. Identify that the feature is rainy weather,
program.

Images