JP2007208802A - Image encoding apparatus and image encoding method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image encoding apparatus capable of executing, in real time, encoding in which prediction is performed in inter-mode and intra-mode and the prediction is performed in the mode of higher encoding efficiency, using software. <P>SOLUTION: An image encoding apparatus 10 is configured by providing, in an intra-coding unit 2, an encoding section 231 which performs orthogonal transformation on an image signal with the start of an image signal inter-mode prediction due to an inter-coding unit 1, starting image signal intra-mode prediction with the decoding end of orthogonal transform coded data resulting from the orthogonal transformation in the coding section, and ending intra-prediction operation with the end of image signal Hadamard transformation due to the inter-coding unit 1. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an image coding apparatus and an image coding method that perform coding by selecting a coding mode with higher coding efficiency from among an inter prediction mode and an intra prediction mode.

Recently, a full-fledged digital video era has come, such as the transmission power of digital terrestrial broadcasting using MPEG-2 encoding being full power.
On the other hand, in the field of communication, H.P. Development using MPEG4-AVC called H.264 has also been made. H. It is preferable that an economical LSI capable of executing the H.264 encoding method in real time can be obtained.

Patent Document 1 discloses an image information encoding apparatus capable of realizing high-speed encoding processing even when vector information necessary for adjacent blocks for parallel processing cannot be obtained because real-time encoding is performed. Has been. That is, in an image information encoding apparatus that outputs image compression information based on an image encoding method such as MPEG4 / AVC, when determining whether or not a predetermined block is in a skip mode or a spatial direct mode. The motion vector information and the like need to be calculated for all of the predetermined adjacent blocks. When parallel processing is performed for each block in order to increase the overall processing speed, not all motion vector information and the like of a predetermined adjacent block are necessarily obtained. In this case, in order to be able to perform mode determination without waiting for calculation of motion vector information etc. for adjacent blocks, mode determination is performed using pseudo motion vector information etc. of other neighboring blocks instead of adjacent blocks. An image information encoding apparatus is disclosed.
JP 2005-70938 A

  However, in the image information encoding device disclosed in Patent Document 1, processing such as motion vector detection, motion compensation, variable length encoding, and deblocking filter, which require a large amount of processing during encoding, is performed in parallel using dedicated hardware. Real-time encoding is performed by processing. In the case of using dedicated hardware, it is difficult to realize a low-cost encoder as compared to performing all processing by software.

  Accordingly, the present invention has been made to solve the above-described problems, and predicts an image in an inter mode and an intra mode, and predicts an image with a higher encoding efficiency. An object is to provide an image encoding apparatus and an image encoding method that can be executed in real time by software.

The first invention in the present invention is an image code for performing main encoding of sequentially supplied images using the results of provisional coding in inter prediction mode and provisional coding in intra prediction mode performed while searching for motion vectors. In the encoding apparatus, an encoding unit that performs main encoding of the current encoding target image that is currently the main encoding target to generate encoded data, and a temporary code in the inter prediction mode for the sequentially supplied images An inter-prediction unit that continuously performs encoding, and is used for provisional encoding in the intra prediction mode of the next encoding target image that is to be encoded next to the current encoding target image by decoding the encoded data A decoding unit that generates a reference image; an intra prediction unit that performs provisional encoding of the next encoding target image using the intra prediction mode using the reference image; and the next code Comparing the generated code amount in the temporary encoding in the inter prediction mode for the target image with the generated code amount in the temporary encoding in the intra prediction mode for the same image, the prediction mode with the smaller generated code amount A selection unit that selects the inter-prediction mode, and when the inter-prediction mode is selected by the selection unit, the next code according to the inter-prediction mode using the motion vector obtained by searching by provisional encoding in the inter-prediction mode The encoding unit performs main encoding of the encoding target image, and when the selection unit selects the intra prediction mode, the encoding unit performs main encoding of the next encoding target image in the intra prediction mode. The first control unit to be performed and the inter prediction unit to the encoding unit during the period in which the inter-prediction unit performs temporary encoding on the next encoding target image The next generation using the reference image generated by the intra prediction unit by generating the encoded data for the target image, generating the reference image for the encoded data generated by the decoding unit. Control is performed to perform provisional encoding on the encoding target image, and the main encoding operation in the encoding unit, the temporary encoding operation by the intra prediction unit, and the reference image generation by the decoding unit for the sequentially supplied images An image encoding apparatus comprising: a second control unit that sequentially performs operations.
According to a second aspect of the present invention, there is provided an image encoding method for performing main encoding of sequentially supplied images using a result of provisional encoding in inter prediction mode and provisional encoding in intra prediction mode performed while searching for a motion vector. An encoding step of performing main encoding of a current encoding target image that is a current encoding target to generate encoded data, and temporary encoding in the inter prediction mode for the sequentially supplied images. An inter-prediction step, and decoding of the encoded data to generate a reference image used in provisional encoding of the next encoding target image to be encoded next to the current encoding target image in the intra prediction mode A decoding step, an intra prediction step of performing temporary encoding of the next encoding target image using the intra prediction mode using the reference image, Compare the generated code amount in the temporary encoding in the inter prediction mode with respect to the encoding target image and the generated code amount in the temporary encoding in the intra prediction mode with respect to the image, and A selection step of selecting a prediction mode, and if the inter prediction mode is selected in the selection step, search for the main encoding of the next encoding target image in the encoding step by provisional encoding in the inter prediction mode When the intra prediction mode is selected in the selection step, the encoding of the next encoding target image in the encoding step is performed as the intra prediction. A first control step for controlling the encoding to be performed according to the mode, and the next encoding target image in the inter prediction step The encoding step for generating the encoded data for the current encoding target image during a period of encoding, and a reference image used for provisional encoding of the next encoding target image in the intra prediction mode And a second control step for performing control so as to perform the decoding step for generating the image and the intra prediction step for performing the temporary encoding of the next encoding target image in the intra prediction mode. An image encoding method is provided that performs the main encoding of the image.

  According to the present invention, the encoding unit that generates the encoded data by performing the main encoding of the current encoding target image that is currently the main encoding target, and the provisional encoding in the inter prediction mode for the sequentially supplied images An inter prediction unit that continuously performs encoding, and a reference that is used in provisional encoding in the intra prediction mode of the next encoding target image that is to be encoded next to the current encoding target image by decoding the encoded data A decoding unit that generates an image; an intra prediction unit that performs provisional encoding of the next encoding target image in the intra prediction mode using the reference image; and a temporary code in the inter prediction mode for the next encoding target image And the generated code amount in the provisional encoding in the intra prediction mode for the same image is compared, and the prediction mode with the smaller generated code amount is selected. When the inter prediction mode is selected by the selection unit and the selection unit, the motion vector obtained by searching by provisional encoding in the inter prediction mode is used to determine the next encoding target image in the inter prediction mode. First encoding is performed by the encoding unit, and when the intra prediction mode is selected by the selection unit, the encoding unit performs main encoding of the next encoding target image in the intra prediction mode. 1, in a period in which the inter prediction unit performs temporary encoding on the next encoding target image, the encoding unit generates the encoded data for the current encoding target image, The decoding unit generates the reference image for the encoded data, and the intra prediction unit uses the reference image generated for the next encoding target image. And sequentially performing the main encoding operation in the encoding unit, the temporary encoding operation by the intra prediction unit, and the reference image generation operation by the decoding unit for the sequentially supplied images. Since there is a special configuration provided with a second control unit to be performed, the image signal encoding is performed by software by predicting in the inter mode and the intra mode and predicting in the mode with the higher encoding efficiency. An image encoding device and an image encoding method that can be executed in time can be realized.

Hereinafter, an image coding apparatus according to an embodiment of the present invention will be described with reference to FIGS.
FIG. 1 is a block diagram illustrating a configuration example of an image encoding device according to an embodiment of the present invention. FIG. 2 is a diagram illustrating an operation example of the image coding apparatus according to the embodiment of the present invention. FIG. 3 is a diagram illustrating a configuration example of a main part of the image encoding device according to the embodiment of the present invention. FIG. 4 is a diagram (No. 1) for explaining an operation example of the image coding device according to the embodiment of the present invention. FIG. 5 is a diagram (No. 2) for explaining an operation example of the image coding device according to the embodiment of the present invention.

  The image encoding device is currently used for the purpose of realizing real-time encoding of an image signal that is predicted in inter mode and intra mode and predicted in the mode with higher encoding efficiency by software execution. An encoding unit that performs the main encoding of the current current encoding target image and generates encoded data; an inter prediction unit that continuously performs temporary encoding in inter prediction mode for sequentially supplied images; A decoding unit that decodes the encoded data and generates a reference image used in provisional encoding in an intra prediction mode of a next encoding target image to be encoded next to the current encoding target image; and the reference An intra prediction unit that performs temporary encoding of the next encoding target image in the intra prediction mode using an image, and the inter prediction mode for the next encoding target image. A selection unit that selects a prediction mode with a smaller amount of generated code by comparing a generated code amount in temporary encoding with a code and a generated code amount in temporary encoding with the intra prediction mode for the same image; When the inter prediction mode is selected by the selection unit, main encoding of the next encoding target image in the inter prediction mode using a motion vector obtained by searching by provisional encoding in the inter prediction mode. First control for causing the encoding unit to perform main encoding of the next encoding target image in the intra prediction mode when the intra prediction mode is selected by the selection unit. Generating the encoded data for the current encoding target image in the encoding unit during a period in which the inter prediction unit performs temporary encoding for the next encoding target image. Causing the decoding unit to generate the reference image with respect to the encoded data, and the intra prediction unit to perform temporary encoding on the next encoding target image using the reference image. Second control for sequentially performing a main encoding operation in the encoding unit, a temporary encoding operation by the intra prediction unit, and a reference image generation operation by the decoding unit for the sequentially supplied images. This is realized by providing a part.

The configuration of the image encoding device will be described.
The image encoding device 10 shown in FIG. 1 includes an inter encoding unit 1, an intra prediction processing unit 21, a mode determination unit 22, an intra encoding unit 2 including an encoding / decoding unit 23, and a control unit 3. Composed. The control unit 3 stores a CPU (not shown) and controls the operation of each unit of the circuit unit.
The intra prediction processing unit 21 of the image encoding device 10 illustrated in FIG. 3 includes a prediction image generator 211, a selector 212, a residual calculator 213, a Hadamard transformer 214, an absolute value summation device 215, a SATD (Sum of Absolute Transform). Differences) evaluator 216 and prediction mode indicator 217. The encoding / decoding unit 23 includes an orthogonal transformer 231, a quantizer 232, an inverse quantizer 233, an inverse orthogonal transformer 234, and an image memory unit 235. The image memory unit 235 stores the decoded image.

The operation of the image encoding device 10 will be described.
First, the processing of the first frame will be described.
A macroblock image obtained by dividing horizontal and vertical into 16 pixels × 16 pixels is input to the intra prediction processing unit 21 of the inter coding unit 1 and the intra coding unit 2. The inter coding unit 1 stores the input image in an image memory (frame memory) (not shown) in order to perform motion prediction of the input image input in the next frame. The intra prediction processing unit 21 divides the macroblock image into 16 4 × 4 pixel blocks, and predicts from the pixel blocks adjacent to the left and upper sides of the macroblock image while predicting 4 × 4 pixels. I do. Similarly, intra prediction temporary encoding of an image divided into four 8 × 8 pixel blocks and intra prediction temporary encoding of an image of 16 × 16 pixel blocks are performed. In the intra prediction provisional encoding, orthogonal transformation of each pixel block is performed by Hadamard transform with simple calculation instead of DCT (discrete cosine transform). The transform time of Hadamard transform can be calculated in a short time compared to orthogonal transform by DCT. An evaluation result for evaluating the encoding method can be easily obtained.

  The mode determination unit 22 performs intra prediction temporary encoding of a 4 × 4 pixel block image, intra prediction temporary encoding of a divided image of 8 × 8 pixel block, and intra prediction temporary encoding of a 16 × 16 pixel block image. The generated code amounts obtained are compared. Intra prediction provisional encoding that gives the smallest generated code amount is determined. The encoding / decoding unit 23 divides the input macroblock image into pixel blocks and performs intra prediction encoding using the result of the intra prediction provisional encoding determined in the mode. The intra-prediction coding is normal intra-prediction coding in which orthogonal transformation is performed by DCT (hereinafter, normal coding may be referred to as main coding). Predictive encoded data obtained by the main encoding is supplied to an entropy encoding unit (not shown) and subjected to variable length encoding. The encoding / decoding unit 23 decodes the prediction encoded data, and the decoded image of the obtained macroblock is supplied to the intra prediction processing unit 21 as a reference image for intra prediction. The decoded image is supplied to a deblocking filter (not shown). Further, the decoded image is stored at a predetermined macro block position in the frame memory as a reference image used for motion vector search in the next frame and subsequent frames.

Next, the second and subsequent frames will be described.
The inter coding unit 1 performs motion prediction with reference to frame image data that has been encoded and decoded and stored. In motion prediction, the position of a reference image similar to an input block image is obtained by searching. The motion vector is obtained from the relationship of the position of the reference image that gives the smallest mean square error. This is a process with a large amount of calculation. The input image is provisionally encoded using the obtained motion vector. The temporary encoding is obtained by Hadamard transform of the difference between the input image and the reference image. In the same way as described above, the intra coding unit 2 performs intra prediction provisional coding of a 4 × 4 pixel block image, intra prediction provisional coding of a divided image of 8 × 8 pixel block, and pixel block of a 16 × 16 pixel block image. Intra-prediction provisional encoding is performed for each.
The mode determination unit 22 generates the generated code amount obtained by the provisional encoding performed by the inter encoding unit 1, the 4 × 4 pixel, the 8 × 8 pixel obtained by the provisional encoding performed by the intra encoding unit 2, And the generated code amount in the case of division into 16 × 16 pixel blocks, the coding mode that gives the smallest generated code amount is determined. The encoding / decoding unit 23 performs the main encoding and decoding of the input image by a normal method according to the determined encoding mode. When encoding by inter prediction is determined, the motion vector obtained by searching by provisional encoding is used to perform main encoding. When coding by intra prediction is determined, main coding is performed by block division that gives the smallest generated code amount among block divisions of 4 × 4 pixels, 8 × 8 pixels, and 16 × 16 pixels. The

This will be further described with reference to FIG.
FIG. 4A shows the time for performing inter prediction encoding, and FIG. 4B shows the time for intra prediction encoding and encoding / decoding. (N) indicates that the order of macroblocks for predictive coding is the nth time.
Inter prediction in (A) is performed in succession by inter prediction (n-1), inter prediction (n), and inter prediction (n + 1). In these inter predictions, only the provisional encoding described above is performed. The processing in (B) includes intra prediction (n−1), coding / decoding (n−1), intra prediction (n), coding / decoding (n), intra prediction (n + 1), coding / decoding ( n + 1). The temporary encoding, the main encoding, and the decoding are repeatedly performed. That is, the inter prediction (n−1) started from time t1 ends at time t3, and inter prediction (n) is performed immediately thereafter. Intra prediction (n-1) starts at t2 and ends at t3. From t3 to t4, encoding / decoding (n-1) is performed.

Regarding the n-th process, inter prediction is started from t3, and intra prediction is started from t4. The provisional encoding of both the inter prediction and the intra prediction ended at t5 ends. At that time, the mode determination unit determines which of the provisional encodings the generated code amount is small. The main encoding and decoding are performed during the period from t5 to t6. Even when it is determined that encoding is performed by inter prediction, normal inter prediction encoding and decoding are performed in a period from t5 to t6 in (B). Inter coding (n + 1) for searching for a motion vector for a macroblock image using a frame image generated by performing main coding and decoding in the past as a frame unit as a reference image is immediately after inter coding (n). You can start at t5.
When inter coding is performed by hardware, a circuit that performs inter prediction temporary coding, normal inter prediction coding, and decoding is arranged in a circuit unit that performs hardware inter coding. Therefore, the inter prediction (n) is performed in the period from t4 to t5. On the other hand, in the case of the present image encoding apparatus 10, since regular inter prediction encoding and decoding are performed using the function of the intra encoding unit, the inter prediction temporary encoding of (n) is performed at t3 to t5. It can be executed in a period.

Next, this will be described in detail.
The intra prediction processing unit 21 and the encoding / decoding unit 23 will be described with reference to FIG.
First, the intra prediction process will be described.
First, the predicted image generator 211 generates a predicted image to be used for intra prediction based on past encoded and decoded image data stored in the image memory unit 235 of the encoding / decoding unit 23. The selector 212 selects and outputs a signal input to the a side in FIG. The residual calculator 213 compares the input image with the predicted image and calculates a residual. The Hadamard transformer 214 simply orthogonally transforms the residual. The absolute value summer 215 obtains the sum of absolute values of coefficient values obtained by Hadamard transform. The SATD evaluator 216 evaluates the sum of absolute values of coefficient values obtained by block division of 4 × 4 pixels, 8 × 8 pixels, and 16 × 16 pixels, and gives the sum of the smallest absolute values. Outputs the sum of absolute value of block division number and coefficient value.

The encoding / decoding unit 23 will be described.
The encoding / decoding unit 23 generates and encodes and decodes a reference image used for intra prediction at the time of intra prediction, and performs main encoding and decoding according to the mode determined by the mode determination unit 22.
The processing at the time of intra prediction will be described.
The orthogonal transformer 231 performs normal orthogonal transformation of the residual data generated by the residual calculator 213, that is, DCT, and obtains coefficient data. The quantizer 232 quantizes the coefficient data with a predetermined quantization width (quantization step) to generate quantized data. The inverse quantizer 233 dequantizes the quantized data to obtain coefficient data. The inverse orthogonal transformer 234 performs inverse DCT transform on the coefficient data to obtain decoded residual data. The image memory unit 235 acquires the predicted image (reference image) generated by the predicted image generator 211 via the selector 212, and generates decoded image data from the predicted image and decoded residual data. The decoded image data is supplied to the predicted image generator 211, where it is used as a predicted image for use in subsequent encoding.

A process when it is determined that the mode determination unit 22 performs intra coding will be described.
The residual calculator 213 of the intra prediction processing unit 21 compares the input image for performing the main encoding with the prediction image for main encoding generated by the prediction image generator 211 to obtain residual data. The orthogonal transformer 231 of the encoding / decoding unit 23 performs DCT conversion of the residual data. The quantizer 232 quantizes the residual data and outputs the obtained quantized data as predictive encoded data to an entropy encoder (not shown).

Processing when the mode determination unit 22 determines that inter coding is to be performed will be described.
A predicted image (reference image) generated using the motion vector obtained at the time of provisional encoding in the inter encoding unit 1 is input to the b side of the selector 212. The selector 212 selects data input to the b side based on the determination result of the mode determination unit 22 supplied via the prediction mode indicator 217. The residual calculator 213 compares the input image with the predicted image and generates residual data. The orthogonal transformer 231 obtains coefficient data by DCT transforming the residual data, and the quantizer 232 quantizes the coefficient data and outputs the obtained quantized data as predictive encoded data. The inverse quantizer 233 and the inverse orthogonal transformer 234 dequantize and inverse DCT the prediction encoded data to obtain decoded residual data. The image memory unit 235 generates decoded image data based on the predicted image input through the selector 212 and the decoded residual data, and stores the decoded image data in an image memory provided therein. The decoded image data is supplied to a deblocking filter (not shown) for attenuating block noise generated during encoding and decoding. The predicted image generator 211 uses the decoded image data to generate predicted image data (reference image data) of an input image to be encoded next time.

The predicted image will be described with reference to FIG.
In the figure, (n) is a 16 × 16 pixel macroblock image to be encoded. (N−1) is a macroblock image encoded immediately before, and (n + 1) is a macroblock image to be encoded next. (M−1) to (m + 1) are macroblock images that have been main-encoded and decoded in the past and arranged adjacent to the upper side of the macroblocks (n−1) to (n + 1).
The image data used when encoding the macroblock image (n) is 16 pixels on the right side of (n−1), 1 pixel on the lower right side of (m−1), and 16 pixels on the lower side of (m). Pixels and 16 pixels below (m + 1). The predicted image generator 211 uses the decoded image data obtained in the past by the encoding / decoding unit 23 to generate pixel data as these predicted images.

The prediction image used at the time of intra prediction encoding is described with reference to FIG.
In the figure, (n) is a macroblock to be encoded, as shown in FIG. Intra-prediction encoding is predictive encoding with 16 4 × 4 pixel blocks, 4 8 × 8 pixel blocks, or 1 16 × 16 pixel block. The 4 × 4 pixel block indicated by diagonal lines in the figure is subjected to predictive encoding using the data of the pixels with nine halftone dots shown on the left and upper sides thereof. The encoding / decoding unit 23 performs the main encoding and decoding at the time of the temporary encoding of the macroblock (n) to generate decoded image data for generating a predicted image.
The main encoding and decoding performed by the encoding / decoding unit 23 is for predictive image generation at the time of intra prediction encoding, main encoding and decoding when determined to be intra prediction encoding, and further inter prediction. The main encoding and decoding when it is determined to be encoding is executed. Therefore, the inter encoding unit 1 only needs to perform inter provisional encoding for mode determination. The orthogonal transformation also uses Hadamard transformation. The intra coding unit 2 performs the latter half of the calculation at the time of inter coding. Since the inter encoding unit 1 only needs to mainly perform motion vector calculations related to the first half of the calculation, real-time encoding calculations can be performed by software processing.

  As described above, according to the image encoding device 10 shown in the present embodiment, the encoding unit that performs the main encoding of the current encoding target image that is the current encoding target and generates encoded data. 231 and 232, an inter prediction unit 1 that continuously performs provisional encoding of sequentially supplied images in an inter prediction mode, and an encoding target next to the current encoding target image by decoding the encoded data Decoding units 233 and 234 that generate reference images to be used in provisional encoding in the intra prediction mode of the next encoding target image, and provisional encoding in the intra prediction mode of the next encoding target image using the reference image An intra-prediction unit 21 for performing encoding, a generated code amount in provisional encoding in the inter prediction mode for the next encoding target image, and a provisional code in the intra prediction mode for the same image And a selection unit 22 that selects the prediction mode with the smaller generation code amount, and the inter prediction mode when the selection unit selects the inter prediction mode. Using the motion vector obtained by encoding, the encoding unit performs the main encoding of the next encoding target image in the inter prediction mode, and the selection unit selects the intra prediction mode. In this case, a first control unit 3 that causes the encoding unit to perform main encoding of the next encoding target image in the intra prediction mode, and the inter prediction unit 1 performs provisional encoding on the next encoding target image. The encoding units 231 and 232 generate the encoded data for the current encoding target image during a period in which the encoding units 231 and 232 generate the encoded data generated by the decoding units 233 and 234. The reference image is generated with respect to the image, and the intra-prediction unit 21 is controlled to perform provisional encoding on the next encoding target image using the reference image generated. A second control unit 3 for sequentially performing the main encoding operation in the encoding units 231 and 232, the temporary encoding operation by the intra prediction unit 21 and the reference image generation operation by the decoding units 231 and 232; Since there is a special configuration, an image encoding device that can perform image signal encoding in real time by software predicting in inter mode and intra mode and predicting in the mode with higher encoding efficiency realizable.

It is a block diagram which shows the structural example of the image coding apparatus which concerns on implementation of this invention. It is a figure which shows the operation example of the image coding apparatus which concerns on implementation of this invention. It is a figure which shows the structural example of the principal part of the image coding apparatus which concerns on implementation of this invention. It is FIG. (1) for demonstrating the operation example of the image coding apparatus which concerns on implementation of this invention. It is FIG. (2) for demonstrating the operation example of the image coding apparatus which concerns on implementation of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Inter encoding part 2 Intra encoding part 21 Intra prediction process part 22 Mode determination part 23 Encoding / decoding part 211 Predictive image generator 212 Selector 213 Residual calculator 214 Hadamard transformer 215 Absolute value summation device 216 SATD Evaluator 217 Prediction mode indicator 231 Orthogonal transformer 232 Quantizer 233 Inverse quantizer 234 Inverse orthogonal transformer 235 Image memory unit

Claims (2)

In an image encoding device that performs main encoding of sequentially supplied images using the results of provisional encoding in inter prediction mode performed while searching for motion vectors and provisional encoding in intra prediction mode,
An encoding unit that performs main encoding of the current encoding target image that is currently the main encoding target and generates encoded data;
An inter prediction unit that continuously performs temporary encoding in the inter prediction mode for the sequentially supplied images;
A decoding unit that decodes the encoded data and generates a reference image to be used for provisional encoding in the intra prediction mode of a next encoding target image to be encoded next to the current encoding target image;
An intra prediction unit that performs provisional encoding of the next encoding target image in the intra prediction mode using the reference image;
The generated code amount is small by comparing the generated code amount in the temporary encoding in the inter prediction mode for the next encoding target image with the generated code amount in the temporary encoding in the intra prediction mode for the same image. A selection unit for selecting a prediction mode of
When the selection unit selects the inter prediction mode, the encoding of the next encoding target image in the inter prediction mode is performed using the motion vector obtained by searching by provisional encoding in the inter prediction mode. First control for causing the encoding unit to perform main encoding of the next encoding target image in the intra prediction mode when the intra prediction mode is selected by the selection unit. And
During the period when the inter prediction unit performs provisional encoding on the next encoding target image, the encoding unit generates the encoded data for the current encoding target image, and the decoding unit generates the encoded data. The reference image is generated for the encoded data, and the intra prediction unit is controlled to perform provisional encoding on the next encoding target image using the reference image generated. A second control unit that sequentially performs a main encoding operation in the encoding unit, a provisional encoding operation by the intra prediction unit, and a reference image generation operation by the decoding unit,
An image encoding apparatus comprising: In an image encoding method for performing main encoding of sequentially supplied images using the results of temporary encoding in inter prediction mode and temporary encoding in intra prediction mode performed while searching for a motion vector,
An encoding step of performing main encoding of a current encoding target image that is currently an encoding target to generate encoded data;
An inter prediction step for continuously performing temporary encoding in the inter prediction mode for the sequentially supplied images;
A decoding step of decoding the encoded data and generating a reference image to be used for provisional encoding in the intra prediction mode of the next encoding target image to be encoded next to the current encoding target image;
An intra prediction step of performing temporary encoding in the intra prediction mode of the next encoding target image using the reference image;
The generated code amount is small by comparing the generated code amount in the temporary encoding in the inter prediction mode for the next encoding target image with the generated code amount in the temporary encoding in the intra prediction mode for the same image. A selection step to select a prediction mode of
When the inter prediction mode is selected in the selection step, the motion vector obtained by searching the main encoding of the next encoding target image in the encoding step by temporary encoding in the inter prediction mode is used. If the intra prediction mode is selected in the selection step, control is performed so that the main encoding of the next encoding target image in the encoding step is performed in the intra prediction mode. A first control step,
The encoding step for generating the encoded data for the current encoding target image during a period in which temporary encoding is performed on the next encoding target image in the inter prediction step; and the next encoding target image The decoding step for generating a reference image used for provisional encoding in the intra prediction mode and the intra prediction step for performing provisional encoding of the next encoding target image in the intra prediction mode. A second control step to control;
And a main encoding of the image.

Images