KR20120046836A - Image pre-processor for vehicle and image pre-process method using the same - Google Patents

Abstract

PURPOSE: An image preprocessor and an image preprocessing method using the same are provided to generate an image of a reduced size and to maintain an image angle of an input image. CONSTITUTION: An image input unit(100) receives an image which is photographed through single lens from a single camera(200). An image reducing unit(120) performs sub sampling at set reduction rate. The image reducing unit generates an image which the size is reduced. An image cutting unit(130) crops the input image at set cutting rate.

Description

Image pre-processor for vehicle and image pre-process method using the same}

The present invention relates to a vehicle image preprocessor and an image preprocessing method using the same. More particularly, the present invention relates to an apparatus and a method for generating a plurality of images suitable for a purpose to be used as an input image.

BACKGROUND ART As the application of vehicle image recognition devices, such as a lane departure warning device (LDWS) and a preceding vehicle collision warning device (FCWS), has been expanded, the prevalence of vehicle image storage devices (image accident recording devices) has also increased.

The vehicle image recognition device and the image storage device are mainly installed in the vicinity of the room mirror to look at the front of the vehicle, and although they have structural similarities such as the application of a DSP-based CPU for analyzing or compressing images, they have different view angles. There is a significant difference.

Since the vehicle image recognition apparatus needs to recognize lanes and objects on the road ahead of the vehicle up to 150m, a narrow angle of view (included angle) lens is used, and the range is about 40 ° to 50 °.

On the other hand, a vehicle image storage device (image accident recorder) uses a wide angle lens because it needs to store a situation near the vehicle, and a range of about 90 ° to 120 ° is known to be appropriate.

 FIG. 1A illustrates an image of 100 ° angle of view using a wide angle lens used in a vehicle image storage device, and FIG. 1B illustrates an image of 45 ° angle of view using a narrow angle lens used in a vehicle image recognition device.

In this case, in the case of the image of FIG. 1A, a wider scene than the image of FIG. 1B is contained on one screen, and thus, the image of FIG.

As such, when the vehicle image recognition apparatus and the image storage apparatus want to implement both functions in one system due to the difference in the optimum angle of view, there is a problem in that the performance of the other side is deteriorated when one side of view is applied.

The present invention has been made to solve such a conventional problem, and aims to generate a plurality of images suitable for each vehicle image recognition device and image storage device from one image.

The vehicle image preprocessor according to the present invention for solving the above problems includes an image input unit, an image reduction unit, and an image cutting unit.

The image input unit receives and stores an image captured by a single lens from an external single camera, and the image reduction unit subsamples the input image at a preset reduction ratio to generate an image reduced in size by a reduction ratio. The image cutting unit crops the input image at a preset cutting ratio to generate an image reduced in size by the cutting ratio.

As a result, as a single input image, a wide angle of view of the input image can be maintained and the screen size can be reduced to generate an image for an image storage device, and the image recognition can be reduced by maintaining the resolution of the input image. An image for the device may be generated.

In this case, the reduction ratio and the cutting ratio may be the same.

In addition, the image input unit is taken at the same angle of view as the image to be created in the image reduction unit, which is an inverse multiple of the reduction ratio or the cutting ratio, and the image captured at the same resolution per unit area as the image to be generated in the image cutting unit, the camera. Can be input from and stored.

In addition, the image cutting unit may perform cropping by setting different cutting ratios to the top, bottom, left, and right sides of the input image according to the position of the image to be generated.

The image reduction unit and the image cutting unit may simultaneously perform subsampling and cropping on the input image when the captured image is input to the image input unit.

Also, when the captured image is input to the image input unit, the image reduction unit and the image cutting unit may perform subsampling and cropping on the input image according to a user's selection.

In addition, the vehicle image preprocessor according to the present invention may further include an image storage and output unit for storing the image generated by the image reduction unit and the image cutting unit, and outputs at the request of the user.

In the vehicle image preprocessing method according to the present invention, a step of receiving and storing an image photographed through a single lens from an external single camera, subsampling the input image at a preset reduction ratio, (B) generating a reduced image, and (c) cropping the input image at a preset cutting ratio to generate an image reduced in size by the cutting ratio.

In this case, the reduction ratio and the cutting ratio may be the same.

In addition, in step (a), the image is taken at the same angle of view as the image to be created in step (b) with the reduction ratio or the inverse multiple of the cutting ratio, and the same resolution per unit area as the image to be created in step (c). The received image from the camera can be stored.

In addition, in step (c), cropping may be performed by setting cutting ratios differently on the top, bottom, left, and right sides of the input image according to the position of the image to be generated.

In addition, in steps (b) and (c), when the image photographed in step (a) is input, subsampling and cropping may be simultaneously performed on the input image.

In addition, in steps (b) and (c), when the image photographed in step (a) is input, subsampling and cropping may be performed on the input image according to a user's selection.

In addition, the vehicle image preprocessing method according to the present invention, after the step (c), further comprising the step of (d) storing the image generated in the step (b) and the step (c), and outputs at the request of the user can do.

An image recognition and image storage device for a vehicle according to the present invention includes a vehicle image preprocessor having the above configuration; An image storage unit for storing an image generated by the image reduction unit; And an image recognizing unit recognizing at least one of a lane and an object on a road from the image generated by the image cutting unit.

By performing subsampling and cropping on a single image input from the outside according to the present invention, an image of reducing the size of the screen while maintaining the angle of view of the input image and an image of reducing the size of the screen while maintaining the resolution of the input image By generating the respective, it is possible to integrate the functions of the vehicle image recognition device and the image storage device into one system.

In particular, the present invention is an image recognition device for a vehicle by generating an image taken by a single camera to maintain the angle of view of the input image and to reduce the size of the screen, and to maintain the resolution of the input image and to reduce the size of the screen In addition, by integrating a single image storage device, the number of systems installed in a vehicle can be reduced, thereby reducing costs for vehicle convenience / safety equipment.

In addition, the present invention is to perform a sub-sampling and cropping from a single image taken by a single wide-angle camera to shoot a wide range of screens, to maintain the angle of view of the input image and to reduce the size of the screen, and the input image By maintaining the resolution of the image and reducing the size of the screen, respectively, the image required for the vehicle image recognition device and the image storage device can be generated using a single camera.

Figure 1 (a) is a view showing an image of 100 ° angle of view using a wide-angle lens used in the vehicle image storage device.
Figure 1 (b) is a view showing an image of 45 ° angle of view using a narrow angle lens used in the vehicle image recognition device.
Figure 2 is a block diagram schematically showing the configuration of a vehicle image preprocessor according to the present invention.
3 is a diagram illustrating a process of generating a storage image from an input image through 2: 1 subsampling;
4 is a diagram illustrating a process of generating an image for recognition from an input image through 2: 1 cropping.
FIG. 5 is a diagram illustrating a vehicle image preprocessor of FIG. 2 as a logic circuit. FIG.
6 is a flowchart illustrating an embodiment of a vehicle image preprocessing method according to the present invention.
7 is a flowchart illustrating still another embodiment of a vehicle image preprocessing method according to the present invention;

Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings. In order to more clearly understand the present invention, the same reference numerals are used for the same components in different drawings.

2 is a block diagram schematically showing the configuration of the vehicle image preprocessor 100 according to the present invention.

The vehicle image preprocessor 100 of the present invention converts the input image into a form suitable for performing a predetermined purpose. Specifically, the vehicle image preprocessor 100 is suitable for checking whether the input image is a lane departure and whether a preceding vehicle collides. It is converted into a form (a form suitable for image recognition), and a form (a form suitable for storing images) for recording in case of a vehicle accident.

The image preprocessor 100 includes an image input unit 110, an image reduction unit 120, an image cutting unit 130, and an image storage and output unit 140.

The image input unit 110 receives an image (an image of the front of the vehicle) photographed through a single lens from an external single camera 200, and the input image may be an image having a larger size than the image to be finally generated.

The digital image is composed of pixels (pixels) representing a point on the screen, and the resolution indicates how many pixels are used to represent the digital image.

The size of the image is determined according to the number of pixels arranged horizontally and vertically, and the resolution may be expressed by multiplying the number of horizontal pixels by the number of vertical pixels. That is, the resolution of the image may be used in combination with the size of the image.

In addition, at the same resolution, the smaller image is clearer, and the sharper image is reduced as the larger image is larger because the same number of pixels are distributed while the area is larger. In this case, the size of the pixel is larger than that of a small image.

In other words, when the size of the image is fixed, the higher the number of pixels, the higher resolution image can be represented.

These resolutions are standardized or customarily used, and the following represent the resolutions generally used.

VGA (Video Graphic Array, 640 × 480): The native resolution of digital video

QVGA (Quarter VGA, 320 × 240): VGA / 4

SVGA (Super VGA, 1280 × 960): VGA × 4

-D1 (720 × 480): Standard resolution of NTSC video

-SD (1280 × 720): Initial HD resolution

Full HD (1920 × 1080): HD resolution

The resolution of such an image should be understood in connection with the angle of view when viewing the camera.

In other words, if the angle of view of the screen is different even though the size of the image is the same, the resolution can be said to be different based on the same object in the image. The resolution is high because there are many pixels. On the contrary, the larger the angle of view of the screen is, the smaller the pixels are for implementing the object.

In the present invention, the image to be finally generated is used to record a high resolution image that can be used for checking lane departure and confirming a preceding vehicle crash (for image recognition), and when a vehicle accident occurs ( Since it is a wide field of view image that can be used for image storage, in order to generate the above-described two kinds of images from one image, the external camera 200 includes all the features of the images to be finally generated. That is, the image is taken by widening the angle of view at a high resolution, and the image input unit 110 receives and stores the image captured by the camera 200.

That is, if the size of the images to be generated (the output image for image storage and image recognition in FIG. 2) is horizontal H and vertical V, the input image includes all the features of the images to be generated. The image is larger than the images to be created at 2 × H and 2 × V, the angle of view is the same wide angle of view as the image storage image, and the resolution is the same high resolution image as the image recognition image to be generated. Would be desirable.

In the present invention, the size of the input image is larger than the size of the images to be generated to be twice as large as the size of the description, but the image may be input by increasing the size to another multiple other than 2 times.

The image reduction unit 120 generates an image having a wide angle of view that can be used (for image storage) to record when a vehicle accident occurs from the input image.

FIG. 3 illustrates a process of generating a storage image from an input image through 2: 1 subsampling.

Since the size of the input image ((a) of FIG. 3) is 2 × H and 2 × V, and the size of the image (FIG. 3 (c)) to be generated is horizontal H and V, the input image Applying 2: 1 subsampling to an image can reduce its size while maintaining a wide field of view of the image.

Subsampling is one of the techniques of reducing images, which reduces the number of samples by performing resampling at regular intervals on current samples (samples). In other words, the pixel is taken while skipping by the reduction magnification.

Referring to (b) of FIG. 3, when subsampling is performed at a ratio of 2: 1, only one pixel is skipped twice as much as the second pixel, the fourth pixel, and the sixth pixel in the horizontal and vertical directions. The size of the input image can be reduced to 1/2 horizontally or vertically.

As described above, the size of the image is determined according to the number of pixels arranged horizontally and vertically, and the resolution of the image may be represented by multiplying the number of horizontal pixels by the number of vertical pixels. In this case, since the input image and the finally generated image have the same contents, and the pixel sizes of the respective images are also the same, it can be seen that the resolution of the image decreases as the size of the input image is reduced.

If the size of the image input from the image input unit 110 is not three times the horizontal and vertical size of the image to be generated, the image reduction unit 120 adjusts the size of the input image through 3: 1 subsampling. It can be reduced to 1/3 horizontally or vertically.

The image cutting unit 130 generates an image that maintains a high resolution that can be used (for image recognition) to check whether the lane is separated from the input image and whether the preceding vehicle collides.

4 is a diagram illustrating a process of generating an image for recognition from an input image through 2: 1 cropping.

Since the size of the input image ((a) of FIG. 4) is 2 × H and 2 × V horizontally, and the size of the image (FIG. 4 (c)) to be created is horizontal H and V, the input image If 2: 1 cropping is performed, the size of the image can be reduced while maintaining the resolution of the image.

Cropping is to make the final version by removing unnecessary elements in order to express the subject more implicitly than when shooting using the original image taken. In other words, it means to cut out unnecessary portions of the image.

Referring to (b) of FIG. 4, when cropping at a ratio of 2: 1, the size of the image can be reduced to 1/2 by taking only necessary portions to cut the top, bottom, left, and right sides of the input image so as to have a narrow angle of view. have.

As described above, the size of the image is determined according to the number of pixels arranged horizontally and vertically, and the resolution of the image can be represented by multiplying the number of horizontal pixels by the number of vertical pixels. In the case of, if only the portion (image of the same position and size) that is the same content as the image generated finally in the input image, it can be seen that there is no change in the resolution of the image.

In addition, as long as the size of the image generated after cropping is half the size of the input image, it is not necessary to equalize the ratio of the top, bottom, left and right cropped portions of the image.

For example, if a necessary portion of the image is a portion shifted downward from the center, the image may be cropped by 30%, 20%, 25%, and 25% of the image.

If the size of the image input from the image input unit 110 is not three times the width and length of the image to be finally generated, the image reduction unit 120 may determine the size of the input image through 3: 1 cropping. You can reduce the size to 1/3 horizontally or vertically.

The image storage and output unit 140 stores the image generated by the image reduction unit 120 and the image cutting unit 130, and outputs the image according to a user's request.

5 is a diagram illustrating a vehicle image preprocessor 100 of FIG. 2 as a logic circuit.

pclk is a pixel clock, a signal indicating the number of pixels that can be taken in one second, and data is a captured image.

In general, when displaying images on the screen, the pixels are sprinkled from the top left to the right, and when a single line is sprayed, the next line is sprinkled from the left side, and hsync is a signal for changing lines, and also displays a single image. The signal that starts from the top left again to show the next image is called vsync.

The image photographed from the camera 200 receives and stores an input buffer 10 corresponding to the image input unit 110 by the above signals.

Since the vertical subsampling block 20 and the horizontal subsampling block 30 correspond to the image reduction unit 120, the vertical subsampling block 20 and the horizontal subsampling block 30 correspond to the image reduction unit 120. An image for image recognition is generated by subsampling at a preset ratio.

In addition, since the vertical crop block 40 and the horizontal crop block 50 correspond to the image cutting unit 130, the input image stored in the input buffer 10 under the control of the timing generator and flow control block 60. Crops at a preset ratio to generate an image for image storage.

In addition, since the output buffer and multiplexer 70 corresponds to the image storage and output unit 140, the output buffer and multiplexer 70 stores the generated images and outputs them to the outside under the control of the timing generator and flow control block 60.

The image preprocessor 100 may also be configured with code that controls direct memory access (DMA) in the CPU in a series of manners. An example of the configuration is not limited to the logic circuit of FIG. 5.

6 is a flowchart illustrating an embodiment of a vehicle image preprocessing method according to the present invention.

First, an image in which the front surface of the vehicle is photographed with a higher resolution and a larger size than the image to be finally generated is input (S100). In this case, if the size of the image to be generated is horizontal H or vertical V, the input image is generated. Since all the features of the images to be included must be included, the size is larger than the images to be created in the horizontal 2 × H and 2 × V, the angle of view is the same wide angle of view as the image storage image to be generated, and the resolution is It may be an image having the same high resolution as the image for image recognition.

When such an image is input, subsampling and cropping are performed on the input image, respectively, to simultaneously generate an image for storing and recognizing an image from a single image of the front of the vehicle while driving the vehicle (S110). ).

In other words, by inputting a 2: 1 subsample of an input image, a wide angle of view is maintained, and an image is reduced in size by 1/2 of a width and a length, and at the same time, a 2: 1 cropping of an input image is performed. Then, the image size is 1/2 of the width and length of the input image, and generates an image of the necessary portion having a narrowly converted angle of view maintaining a high resolution.

The generated images are stored and output according to a user's request (S120), recording when a vehicle accident occurs (using an image generated by subsampling), checking lane departure and checking whether a vehicle crashed in advance (image generated by cropping) Can be used).

7 is a flowchart illustrating still another embodiment of a vehicle image preprocessing method according to the present invention.

As shown in FIG. 6, an image in which the front surface of the vehicle is photographed with a higher resolution and a larger size than the image to be finally generated is input (S200). In this case, if the size of the image to be generated is horizontal H or vertical V, the input is performed. Since the image to be included includes all the features of the image to be generated, the size is larger than the image to be generated in the horizontal 2 × H, 2 × V, the angle of view is the same wide angle of view as the image storage image to be generated, The resolution may be an image having the same high resolution as the image for image recognition to be generated.

Whether the input image is converted into image storage or image recognition is selected by the user (S210).

When the conversion for image storage is selected, the input image may be 2: 1 subsampled to maintain a wide angle of view, and generate an image in which the size of the image is reduced to 1/2 each of the length and width (S220). .

If the conversion for image storage is not selected, that is, if the conversion for image recognition is selected, the input image is cropped 2: 1 so that the size of the image is 1/2 the size of the input image. In operation S230, an image of a required portion having a narrow view angle maintained at high resolution may be narrowly converted.

Images generated by performing subsampling and cropping are stored and output according to a user's request (S240). Images generated by subsampling can be used for recording when a vehicle accident occurs, and are generated by cropping. The image may be used to check whether the lane deviates or whether the preceding vehicle collides.

If an image input from the camera 200 is converted for image storage or image recognition, and the user wants to convert the input image to an image for another use (for image recognition, if the image is converted for image storage), In operation S210, the user may generate a desired image (S250).

On the other hand, the vehicle image preprocessor having the configuration described above and the image storage unit for storing the image generated by the image reduction unit; The vehicle image recognition and image storage device may be configured with an image recognition unit that recognizes at least one of a lane and an object on a road from the image generated by the image cutting unit.

The image storage unit functions as a vehicle image storage device, a so-called vehicle black box, and stores an image for a predetermined time when an event such as a collision or collision of a vehicle occurs or an operation by a user (image storage input) occurs.

The image storage unit recognizes that the vehicle is a collision or the occurrence of the collision (event occurrence) by using an impact sensor for sensing the impact amount.

The image stored by the image storage unit may be an image of a predetermined time before the reference point based on the reference time point at which an event occurs, and an image of a predetermined time after the reference point from the predetermined time before the reference point is to be stored. Can be.

The image recognition unit functions as an image recognition device for a vehicle, a so-called lane departure warning device (LDWS), and a preceding vehicle collision warning device (FCWS). Notify or transmit to the control unit of the vehicle to control the driving state of the lane maintenance, such as maintaining the safety distance.

So far I looked at the center of the preferred embodiment for the present invention. Those skilled in the art will appreciate that the present invention can be implemented in a modified form without departing from the essential features of the present invention. Therefore, the disclosed embodiments should be considered in descriptive sense only and not for purposes of limitation. The scope of the present invention is shown in the claims rather than the foregoing description, and all differences within the scope will be construed as being included in the present invention.

10 input buffer
20 vertical subsampling blocks
30 Horizontal subsampling blocks
40 vertical crop blocks
50 Horizontal crop blocks
60 Timing generator and flow control block
70 Output Buffer And multiplexer
100 Car Video Preprocessor
110 Video Receiver
120 Image Reduction
130 video cuts
140 Video storage and output
200 external cameras

Claims (15)

An image input unit configured to receive and store an image photographed through a single lens from an external single camera;
An image reduction unit for subsampling the input image at a preset reduction ratio to generate an image reduced in size by the reduction ratio; And
An image cutting unit for cropping the input image at a preset cutting ratio to generate an image having a size reduced by the cutting ratio; Vehicle image preprocessor comprising a. The method of claim 1,
And the reduction ratio and the cutting ratio are the same. The method of claim 2,
The image input unit,
The camera captures an image taken at the same angle of view as the image to be generated by the image reduction unit at the reduction ratio or an inverse multiple of the cutting ratio, and the image to be generated at the image cutting unit and the same resolution per unit area. Vehicle image pre-processor, characterized in that the input from the store. The method of claim 1,
The image cutting unit,
According to the position of the image to be generated, the vehicle image pre-processor characterized in that the cropping is performed by setting the cutting ratio differently to the upper, lower, left, right of the input image. The method of claim 1,
The image reduction unit and the image cutting unit,
And when the captured image is input to the image input unit, simultaneously performing subsampling and cropping on the input image. The method of claim 1,
The image reduction unit and the image cutting unit,
And when the photographed image is input to the image input unit, subsampling and cropping are performed on the input image according to a user's selection. The method of claim 1,
An image storage and output unit which stores the image generated by the image reduction unit and the image cutting unit, and outputs the image according to a user's request; Vehicle image preprocessor further comprising a. A vehicle image preprocessor according to any one of claims 1 to 7;
An image storage unit for storing an image generated by the image reduction unit;
And an image recognition unit for recognizing at least one of a lane and an object on a road from the image generated by the image cutting unit. (a) receiving and storing an image photographed through a single lens from an external single camera;
(b) subsampling the input image at a preset reduction ratio to generate an image reduced in size by the reduction ratio; And
(c) cropping the input image at a preset cutting ratio to generate an image having a size reduced by the cutting ratio; Vehicle image preprocessing method comprising a. The method of claim 9,
And the reduction ratio and the cutting ratio are the same. The method of claim 10,
In the step (a),
The image is captured at the same angle of view as the image to be generated in the step (b), and the image captured at the same resolution per unit area as the image to be generated in the step (c). And storing and receiving the input from the camera. The method of claim 9,
In the step (c),
According to the position of the image to be generated, the image pre-processing method for a vehicle characterized in that the cropping is performed by setting the cut ratio differently to the upper, lower, left, right of the input image. The method of claim 9,
In step (b) and step (c),
And in the step (a), when the captured image is input, simultaneously performing subsampling and cropping on the input image. The method of claim 9,
In step (b) and step (c),
When the captured image is input in the step (a), subsampling and cropping are performed on the input image according to a user's selection. The method of claim 9,
After step (c),
(d) storing the image generated in the steps (b) and (c) and outputting the image according to a user's request; Vehicle image pre-processing method comprising a further.

Images