Video processing apparatus and video processing method

Abstract

The present application relates to a video processing apparatus and a video processing method. The video processing apparatus comprises: a scene determination unit which determines a scene of an input image; a similarity determination unit which determines the similarity between frames of the input image using image features; a set information generation unit which calculates, on the basis of the determination result of the scene and the image features of the frame, target set information indicating a set value of video processing; a calculation unit which calculates switching speed information on the basis of the determination result of the similarity, the switching speed information representing a set value switching speed of switching from the set value of the video processing for the input image before scene switching to the set value of the video processing for the input image after scene switching; a set information determination unit which determines variable speed setting information, the variable speed setting information being used for causing the set value of the video processing for the input image to reach the set value based on the target set information at a speed based on the switching speed information.

Description

Video processing device and video processing method

[0001] Cross Reference to Related Applications

[0002] This application claims priority to Japanese Patent Application No. 2024-088314, filed May 30, 2024, and titled “Video processing device and video processing method,” the entire contents of which are incorporated herein by reference. TECHNICAL FIELD

[0003] The present application relates to a video processing device and a video processing method employed in a television display device or the like. BACKGROUND

[0004] In a video processing device such as a television display device, various video processing for improving the quality of a picture is performed. In order to perform video processing corresponding to an input video, analysis of the input video is performed, and video processing of timing and intensity corresponding to the analysis result is implemented.

[0005] For example, with respect to an input video, detection of a pixel value, detection of an edge, analysis of a flatness, degree of noise, and the like are required, and based on these analysis results, various video processing such as contrast, color correction, noise removal, and sharpening processing is performed. In addition, a technology of classifying scenes of an input video and performing video processing suitable for the scenes has been developed.

[0006] However, depending on the situation of the switching of the scenes, sometimes a change in the quality of the picture displayed on the screen due to a change in the setting of the video processing is perceived by the user. When there is such a change in the quality, there is a problem that the user feels that the video is unnatural.

[0007] PRIOR ART DOCUMENTS

[0008] PATENT DOCUMENTS

[0009] Patent Document 1: Japanese Patent Application Publication No. 2005-39786 SUMMARY

[0010] An object of the present application is to provide a video processing device and a video processing method that perform quality setting suitable for each scene, and that can make a change in the quality of the picture on the screen due to a change in the quality setting at the time of the switching of the scenes unperceivable by the user.

[0011] The video processing apparatus of the present application includes a scene determination unit that determines a scene based on an input image, a similarity determination unit that determines similarity between frames included in the input image using image features, a setting information generation unit that calculates target setting information indicating a setting value for video processing based on a result of the determination of the scene and the image features of the frames, a switching speed information calculation unit that calculates switching speed information indicating a speed at which the setting value for video processing of the input image changes from before the determination of the scene to after the determination of the scene based on a result of the determination of the similarity, and a setting information decision unit that decides speed-variable setting information for causing the setting value for video processing of the input image to reach the setting value based on the target setting information at a speed based on the switching speed information. BRIEF DESCRIPTION OF DRAWINGS

[0012] FIG. 1 is a block diagram showing a video processing apparatus according to a first embodiment of the present application;

[0013] FIG. 2 is a block diagram showing one example of a detailed structure of the video processing setting switching control unit 3 and the video processing setting unit 4 in FIG. 1;

[0014] FIG. 3 is a flowchart for explaining the operation of the embodiment;

[0015] FIG. 4 is a graph for explaining a switching speed coefficient;

[0016] FIG. 5 is a graph for explaining a switching speed coefficient;

[0017] FIG. 6 is a graph for explaining a switching speed coefficient;

[0018] FIG. 7 is a graph for explaining speed-variable setting information.

[0019] Explanation of Reference Numerals 1... video processing unit, 2... input video scene determination unit, 3... video processing setting switching control unit, 4... video processing setting unit, 31... feature extraction unit, 32... similarity determination unit, 33, 43, 44... delay unit, 34... switching speed information calculation unit, 41... edge extraction unit, 42... setting information generation unit, 45... setting information decision unit. DETAILED DESCRIPTION

[0020] Hereinafter, the embodiments of the present application will be explained in detail with reference to the drawings.

[0021] (First Embodiment)

[0022] Fig. 1 is a block diagram showing a video processing apparatus of a first embodiment of the present application. The present embodiment prevents a user from noticing a change in picture quality on a screen accompanying a change in video processing by determining a scene of an input video and performing video processing setting suitable for the scene, and adjusting a switching speed of the video processing setting according to a switching condition of the scene.

[0023] The video processing apparatus of Fig. 1 includes a video processing section 1, an input video scene determination section 2, a video processing setting switching control section 3, and a video processing setting section 4. Each section of the video processing apparatus can be configured by a processor using a CPU (Central Processing Unit), a GPU (Graphic Processing Unit), a FPGA (Field Programmable Gate Array), a NPU (Neural Processing Unit), or the like. In addition, each section of the video processing apparatus can be controlled by acting in accordance with a program stored in a memory not shown, or can realize a part or all of the functions by an electronic circuit of hardware.

[0024] An input video is supplied to the video processing section 1, the input video scene determination section 2, the video processing setting switching control section 3, and the video processing setting section 4. The video processing section 1 performs video processing on the input video and outputs an output video. For example, the video processing section 1 performs various video processing such as contrast adjustment processing, color correction processing, noise removal processing, sharpening processing, and the like on the input video. By supplying the output video from the video processing section 1 to a display apparatus not shown, a video based on the output video is displayed on a display screen of the display apparatus. As described later, the input video is subjected to optimal video processing suitable for a scene by the video processing section 1, and a user can enjoy a high-quality video based on the input video.

[0025] The input video scene determination section 2 analyzes the input video (input image) and determines which of a plurality of predetermined scenes each image of the input video is to be classified into. For example, the input video is classified by the input video scene determination section 2 into any one of a plurality of scenes such as a night scene, a moving scene, an animation scene, and the like. The input video scene determination section 2 can also perform scene determination using AI (artificial intelligence), for example. The input video scene determination section 2 can employ various inference models constructed by a CNN (Convolutional Neural Network), a Transformer, a random forest, a support vector machine, or the like, for example. The determination result of the scene by the input video scene determination section 2 is supplied to the video processing setting section 4. Note that the input video scene determination section 2 can perform scene determination for each frame or for each predetermined multi-frame period.

[0026] The video processing setting switching control section 3 is supplied with the input video and obtains switching speed information for controlling the speed of switching of the video processing setting at the time of setting switching of the video processing of the input video. The video processing setting switching control section 3 outputs the obtained switching speed information to the video processing setting section 4. Note that the switching speed information indicates the speed of change of the setting value of the video processing, and the video processing setting switching control section 3 can output, for example, a speed coefficient that specifies the speed of setting processing to the video processing setting section 4 as the switching speed information. The video processing setting switching control section 3 obtains the switching speed information at a timing corresponding to the timing of the scene determination by the input video scene determination section 2. Note that the video processing setting switching control section 3 can obtain the switching speed information at a predetermined period such as every frame, for example.

[0027] The video processing setting section 4 is given the input video and the scene determination result from the input video scene determination section 2. The determination result of the scene from the input video scene determination section 2 changes in the switching of the scene, and therefore the video processing setting section 4 determines the switching of the scene by the change in the determination result of the scene. The video processing setting section 4 obtains setting information (hereinafter, also referred to as speed variable setting information) that specifies the setting values of the various video processing corresponding to the respective scenes, based on the input video and the scene determination result from the input video scene determination section 2. For example, the video processing setting section 4 can also obtain the setting values (parameters) that specify the strength and the like of the various video processing in the video processing section 1, as the setting information. In the present embodiment, the video processing setting section 4 changes the setting values while being based on the switching speed information obtained by the video processing setting switching control section 3, and sets to the video processing section 1. That is, the setting values of the video processing setting of the video processing section 1 change at a speed corresponding to the switching speed information after the start of the switching of the video processing. In this way, the switching (change) of the setting values of the video setting in the video processing section 1 is performed at a speed corresponding to the switching speed information obtained by the video processing setting switching control section 3.

[0028] Fig. 2 is a block diagram showing one example of the detailed structure of the video processing setting switching control section 3 and the video processing setting section 4 in Fig. 1.

[0029] The video processing setting switching control section 3 includes a feature extraction section 31, a similarity determination section 32, a delay section 33, and a switching speed information calculation section 34. In addition, the video processing setting section 4 includes an edge extraction section 41, a setting information generation section 42, delay sections 43 and 44, and a setting information decision section 45.

[0030] The feature extraction section 31 obtains the features of the images of the mutually different frames of the input video, in order to measure the similarity of the mutually different frames (images) of the input video. The feature extraction section 31 can also, for example, perform hue histogram conversion, frequency histogram conversion, brightness histogram conversion, reduction, and the like on the object frames of the input video, thereby obtaining the features of the input video as an index of the features of the input video.

[0031] The output of the feature extraction section 31 is supplied to the similarity determination section 32 and the delay section 33. The delay section 33 supplies the output of the feature extraction section 31 to the similarity determination section 32 after delaying it by a prescribed frame period. In this way, the similarity determination section 32 is supplied with the extraction results of the features of the input images before and after the prescribed frame period. The similarity determination section 32 obtains the similarity of the images before and after the prescribed frame period, which indicates how similar the images before and after the prescribed frame are, based on the comparison of the extraction results of the features before and after the prescribed frame period.

[0032] For example, in a case where the feature extraction using the histogram is performed in the feature extraction section 31, the similarity determination section 32 can determine the similarity based on the square error of each representative value. Also, for example, in a case where the feature extraction using the reduced image is performed in the feature extraction section 31, the similarity determination section 32 can determine the similarity based on the SSIM (Structual SIMilarity), the PSNR (Peak Signal to Noise Ratio), or the like. The similarity determination section 32 outputs the determination result of the similarity, which is found by comparing the image features before and after the prescribed frame period, to the switching speed information calculation section 34.

[0033] The switching speed information calculation section 34 generates the switching speed information that decides the switching speed of the setting of the view processing based on the determination result of the similarity, and outputs it to the setting information decision section 45 of the view processing setting section 4. For example, the switching speed information calculation section 34 can find the switching speed coefficient that indicates the degree of the changing speed of the setting value, and provide it to the setting information decision section 45 as the switching speed information. Note that, as described later, in a case where the setting information decision section 45 sets the switching speed coefficient that indicates the relatively slow switching speed in accordance with the switching speed information, the change of the setting value of the view processing setting will be relatively slowly performed for a relatively long time. Also, on the contrary, in a case where the setting information decision section 45 sets the switching speed coefficient that indicates the relatively fast switching speed in accordance with the switching speed information, the change of the setting value of the view processing setting will be relatively quickly (in a relatively short time) performed.

[0034] In the present embodiment, in a case where the similarity between the images is relatively high, the switching speed information calculation section 34 sets the switching speed to be relatively small. Even in a case of the scene switching, there is a case where the similarity of the images before and after the scene is relatively high. In this case, even if the scene is switched, the change of the image is small, so when the setting value of the view processing is changed in conjunction with the switching of the scene, the user easily perceives the change of the image quality. Therefore, in the present embodiment, the switching speed of the setting value is slowed down. Thus, the change of the image quality is relatively slowly performed, so the user hardly perceives the change of the image quality in conjunction with the change of the setting of the view processing.

[0035] On the contrary, in a case where the similarity between the images is relatively low, the switching speed information calculation section 34 sets the switching speed to be relatively large. In a case where the similarity of the images is relatively low at the time of the switching of the scene, that is, in a case where the change of the image is large, the change of the image quality is not noticeable, so the user hardly perceives the change of the image quality in conjunction with the change of the setting value. Therefore, in this case, the switching speed of the setting value is sped up, so the image quality is stabilized in a short time.

[0036] The image processing setting unit 4 controls the image processing settings corresponding to such similarity. In order to perform image processing corresponding to image features in the image processing unit 1, the edge extraction unit 41 of the image processing setting unit 4 can, for example, extract edges as one of the features of the input image. For example, for each image of the input image, the edge extraction unit 41 extracts flat areas and edge areas based on the difference value and average value with surrounding pixels. Alternatively, the edge extraction unit 41 can also obtain features of the input image through various histogram transformations. The feature extraction results of the input image obtained by the edge extraction unit 41 are supplied to the setting information generation unit 42 and the delay unit 43. The delay unit 43 delays the features of the input image obtained by the edge extraction unit 41 for a predetermined frame period and outputs them to the setting information generation unit 42.

[0037] The setting information generation unit 42 calculates setting information for image processing in the image processing unit 1 based on the features of the input image obtained by the edge extraction unit 41 and the scene determination result supplied from the input image scene determination unit 2. For example, regarding regions where the edge portion of the image is shown according to the feature extraction result, the setting information generation unit 42 generates setting information for enhanced sharpening processing; regarding regions where the flat portion of the image is shown and a certain amount of noise is detected, the setting information generation unit 42 can also generate setting information for enhanced noise removal processing. In addition, based on the feature extraction result obtained from the histogram conversion by the edge extraction unit 41, the setting information generation unit 42 can also generate setting information for contrast correction and color correction corresponding to the overall brightness of the image. Furthermore, the setting information generation unit 42 obtains setting information after adjusting these setting information according to the scene determination result from the input image scene determination unit 2.

[0038] In this embodiment, the setting information generation unit 42 outputs the setting information obtained for the nth frame as target setting information Xn to the setting information determination unit 45. Additionally, the setting information generation unit 42 outputs the setting information obtained for the (n-a)th frame preceding the (a)th frame of the nth frame as past setting information Xn-a to the setting information determination unit 45.

[0039] For example, regarding the nth frame, when the feature quantity obtained by the edge extraction unit 41 is set as feature quantity Pn, and the scene determination result obtained by the input video scene determination unit 2 is set as scene determination result Sn, the target setting information Xn and the past setting information Xn-a generated by the setting information generation unit 42 can be represented by the function F1 in the following equation (1). It should be noted that, regarding the (n-a)th frame, the feature quantity obtained by the edge extraction unit 41 is set as feature quantity Pn-a, and the scene determination result obtained by the input video scene determination unit 2 is set as scene determination result Sn-a. Xn=F1(Pn、Sn) Xn-a=F1(Pn-a、Sn-a)…(1)

[0040] Delay unit 43 delays the output of edge extraction unit 41 by a predetermined frame period and provides it to setting information generation unit 42. Delay unit 44 delays the output of input video scene determination unit 2 by a predetermined frame period and provides it to setting information generation unit 42. Through delay units 43 and 44, the aforementioned feature quantity Pn-a and scene determination result Sn-a can be obtained. Therefore, by using the outputs of delay units 43 and 44, setting information generation unit 42 can obtain past setting information Xn-a. It should be noted that a memory can also be provided in setting information generation unit 42 to hold the target setting information Xn for a frame period. In this case, setting information generation unit 42 can obtain past setting information Xn-a from the memory, and delay units 43 and 44 are not required.

[0041] When the switching speed information calculation unit 34 sets the switching speed coefficient calculated in the nth frame as the switching speed coefficient Vn, in the nth frame, the switching speed coefficient Vn, the target setting information Xn, and the past setting information Xn-a are input into the setting information determination unit 45. The setting information determination unit 45 uses the target setting information Xn, the past setting information Xn-a, and the switching speed coefficient Vn to obtain the setting information after adjusting the switching speed according to the change in the setting value based on the switching speed coefficient Vn, namely, the speed variable setting information Yn for the nth frame. This speed variable setting information Yn is used to specify the setting value for video processing in the video processing unit 1.

[0042] For example, when variable speed setting information Yn, based on a switching speed coefficient Vn representing the delay switching speed, is supplied to the video processing unit 1, the following situation can be considered: until the variable speed setting information Yn changes due to the subsequent scene switching, the setting value determined by the variable speed setting information Yn does not reach the setting value determined by the target setting information Xn. Therefore, the setting information determination unit 45 uses not only the target setting information Xn, the past setting information Xn-a, and the switching speed coefficient Vn, but also the past setting information Yn-a from a frame earlier than the nth frame to determine the variable speed setting information Yn. The setting information determination unit 45 outputs the determined variable speed setting information Yn to the video processing unit 1.

[0043] That is, the variable speed setting information Yn obtained by the setting information determination unit 45 can be expressed by the function F2 in the following equation (2): Yn=F2(Xn,Xn-1,Yn-1,Vn)…(2)

[0044] Next, the operation of this embodiment will be described with reference to Figures 3 to 7. Figure 3 is a flowchart illustrating the operation of the embodiment, and Figures 4 to 6 are graphs illustrating the switching speed coefficient. Figure 7 is a graph illustrating the variable speed setting information.

[0045] In S1 of Figure 3, the feature extraction unit 31 of the image processing setting switching control unit 3 extracts features from the input image. The features extracted by the feature extraction unit 31 are provided to the similarity determination unit 32, and are provided to the similarity determination unit 32 after being delayed by the delay unit 33 for a predetermined frame period. The similarity determination unit 32 determines the similarity between the images before and after the predetermined frame period by comparing the features of the images before and after the predetermined frame period (S2).

[0046] The similarity determination unit 32 provides the similarity determination result to the switching speed information calculation unit 34. The switching speed information calculation unit 34 calculates the switching speed information corresponding to the similarity. For example, the switching speed information calculation unit 34 calculates the switching speed coefficient Vn as the switching speed information (S3). The switching speed information calculation unit 34 provides the calculated switching speed coefficient Vn to the setting information determination unit 45 of the video processing setting unit 4.

[0047] Figures 4-6 show the similarity index on the horizontal axis and the switching speed coefficient Vn on the vertical axis, illustrating the variation of Vn with similarity. The examples in Figures 4-6 all demonstrate that the lower the similarity, the larger the switching speed coefficient Vn, and vice versa. Figure 4 shows the linear variation of Vn with respect to similarity. Figure 5 shows the curvilinear variation of Vn with respect to similarity. Figure 6 shows the discrete variation of Vn with respect to similarity.

[0048] In S11 of Figure 3, the edge extraction unit 41 of the image processing setting unit 4 extracts feature values ​​from the input image. The feature values ​​extracted by the edge extraction unit 41 are provided to the setting information generation unit 42, and are delayed by the delay unit 43 for a predetermined frame period before being provided to the setting information generation unit 42. Meanwhile, the input image scene determination unit 2 determines the scene of the input image according to each predetermined frame period and obtains a determination result. The scene determination result obtained by the input image scene determination unit 2 is provided to the setting information generation unit 42, and is delayed by the delay unit 44 for a predetermined frame period before being provided to the setting information generation unit 42.

[0049] The setting information generation unit 42 generates target setting information Xn as setting information for video processing based on the feature quantity Pn and the scene determination result Sn from the edge extraction unit 41 (S12). The setting information generation unit 42 outputs the target setting information Xn obtained based on the feature quantity Pn and the scene determination result Sn to the setting information determination unit 45. In addition, the setting information generation unit 42 outputs the setting information obtained based on the feature quantity Pn-a and the scene determination result Sn-a from a frame earlier than the nth frame as past setting information Xn-a to the setting information determination unit 45 (S13).

[0050] The setting information determination unit 45 determines the speed variable setting information Yn based on the target setting information Xn, the past setting information Xn-a, the switching speed coefficient Vn, and the past speed variable setting information Yn-a, and outputs it to the video processing unit 1 (S14).

[0051] Figure 7 illustrates the variable speed setting information Yn obtained by the setting information determination unit 45. It should be noted that Figure 7 shows an example where frame a is a 1-frame period. That is, the input video scene determination unit 2 determines the scene for each frame, and the similarity determination unit 32 determines the similarity between the images before and after one frame.

[0052] Now, assuming that for the input video of the (n-1)th frame, the video processing unit 1 performs video processing based on past setting information Xn-1 (past speed variable setting information Yn-1). It should be noted that, as shown in FIG. 7, the correction amount (setting value) for the processing based on the past setting information Xn-1 is set as the correction amount CA0. For the input image of the nth frame, the switching speed information calculation unit 34 outputs the switching speed coefficient Vn, the edge extraction unit 41 outputs the quantity feature Pn, the input video scene determination unit 2 outputs the scene determination result Sn, and the setting information generation unit 42 calculates the target setting information Xn based on the quantity feature Pn of the image of the nth frame and the scene determination result Sn of the image of the nth frame. The correction amount for the processing based on this target setting information Xn is set as the correction amount CA1 shown in FIG. 7. The setting information generation unit 42 also outputs information from the past setting information Xn-1.

[0053] The target setting information Xn, past setting information Xn-1, switching speed coefficient Vn, and past variable speed setting information Yn-1 are also provided to the setting information determination unit 45. The setting information determination unit 45 determines the variable speed setting information Yn for changing the correction amount of the video processing from CA0 to CA1. In this case, as shown in FIG7, the variable speed setting information Yn is information that changes the correction amount at a speed corresponding to the switching speed coefficient Vn. In FIG7, the changes in the correction amount determined by each variable speed setting information Yn corresponding to the five switching speed coefficients Vn are shown by arrows. That is, when the similarity between the image of the (n-1)th frame and the image of the nth frame is low and the switching speed coefficient Vn is large, the variable speed setting information Yn that changes the correction amount to the correction amount CA1 in a relatively short time is supplied to the video processing unit 1. Conversely, when the similarity between the image of the (n-1)th frame and the image of the nth frame is high and the switching speed coefficient Vn is small, the variable speed setting information Yn that changes the correction amount to the correction amount CA1 takes a relatively long time is supplied to the video processing unit 1.

[0054] It should be noted that the example in Figure 7 shows an example where the correction amount determined by the past variable speed information Yn-1 at the time of frame n-1 reaches the correction amount (CA0) based on the past setting information Xn-1. If, at the time of frame n-1, the correction amount determined by the past variable speed information Yn-1 does not reach the correction amount (CA0) determined by the past setting information Xn-1, the setting information determination unit 45 calculates variable speed setting information Yn that takes into account the difference between the correction amount determined by the past variable speed information Yn-1 and the correction determined by the past setting information Xn-1, and outputs it to the video processing unit 1.

[0055] In scene transitions, there may be cases where images change significantly (low similarity) or remain almost unchanged (high similarity). When switching scenes between images with high similarity, the changes in image processing accompanying the scene change make the image quality changes easily noticeable. However, in this embodiment, when the similarity between the images before and after the scene transition is relatively high, the switching speed coefficient Vn is set to be relatively small. This results in a slow change rate of the correction amount (set value) in the image processing accompanying the scene transition, making the image quality change difficult for the user to perceive.

[0056] On the other hand, when scenes switch between images with low similarity, the visual changes are significant, and the resulting image quality changes due to the changes in visual processing during scene switching are difficult to perceive. Therefore, in this embodiment, when the similarity between the images before and after the scene switch is relatively low, the switching speed coefficient Vn is set to be relatively large. This increases the rate of change of the correction amount in the visual processing accompanying the scene switch, allowing the target correction amount to be reached in a short time, thus stabilizing the image quality within a short period. In this way, this embodiment, by adjusting the rate of change of the set value based on the similarity between images, is extremely effective as a countermeasure to obtain a natural visual image where image quality changes are not noticeable.

[0057] Thus, in this embodiment, by determining the scene of the input video and setting video processing appropriate to the scene, and adjusting the switching speed of the video processing settings according to scene changes, it is possible to prevent the user from perceiving changes in image quality caused by changes in video processing. Therefore, the user can appreciate natural video without being affected by scene changes.

[0058] This application is not limited to the above-described embodiments, and various modifications can be made during the implementation phase without departing from its main intent. Furthermore, the above embodiments include technical solutions at various stages, and various technical solutions can be extracted through appropriate combinations of the disclosed multiple structural elements. For example, if even if several structural elements are deleted from all the structural elements shown in the embodiments, the problem described in the "Problem to be Solved by the Invention" column can be solved, and the effects described in the "Effects of the Invention" column can be obtained, then the structure with the deleted structural element can be extracted as an invention.

[0059] Furthermore, in the technologies described herein, the controls and functions, primarily illustrated by flowcharts, can mostly be set by a program. This program can be read and executed by a computer to achieve the aforementioned controls and functions. As a computer program product, this program can be recorded or stored, in whole or in part, on removable media such as floppy disks, CD-ROMs, non-volatile memory, hard disks, and volatile memory. It can be distributed or provided when the product is shipped, or via removable media or communication lines. Users can easily implement the video processing apparatus and video processing method of this embodiment by downloading the program via a communication network and installing it on a computer, or by installing it from a storage medium on a computer.

Claims

1. An image processing apparatus, comprising: The video processing apparatus includes: a scene determination section that determines a scene from an input image; a similarity determination section that determines similarity between frames included in the input image using image features; a setting information generation section that generates target setting information indicating a setting value for video processing based on a result of the determination of the scene and the image features of the frames; a switching speed information calculation section that calculates switching speed information indicating a speed of switching from a setting value for video processing of the input image before a change in the result of the determination of the scene to a setting value for video processing of the input image after the change in the result of the determination of the scene based on a result of the determination of the similarity; and a setting information decision section that decides speed-variable setting information for reaching the setting value based on the target setting information at a speed based on the switching speed information.

2. The video processing apparatus according to claim 1, wherein the setting information decision section decides the speed-variable setting information based on: the target setting information and the speed-variable setting information before the change in the result of the determination of the scene, the target setting information after the change in the result of the determination of the scene, and the switching speed information calculated based on similarity between images before and after the change in the result of the determination of the scene.

3. The video processing apparatus according to claim 1, wherein the higher the similarity, the smaller switching speed information calculated by the switching speed information calculation section, and the lower the similarity, the larger switching speed information calculated by the switching speed information calculation section.

4. A video processing method of a video processing apparatus including a scene determination section, a similarity determination section, a setting information generation section, a switching speed information calculation section, and a setting information decision section, wherein the scene determination section determines a scene from an input image, the similarity determination section determines similarity between frames included in the input image using image features, the setting information generation section generates target setting information indicating a setting value for video processing based on a result of the determination of the scene and the image features of the frames, the switching speed information calculation section calculates switching speed information indicating a speed of switching from a setting value for video processing of the input image before a change in the result of the determination of the scene to a setting value for video processing of the input image after the change in the result of the determination of the scene based on a result of the determination of the similarity, and the setting information decision section decides speed-variable setting information for reaching the setting value based on the target setting information at a speed based on the switching speed information. ​ ​

Images