CN103647911A - Image seamless splicing method based on black-level adjustment - Google Patents

Abstract

The invention discloses an image seamless splicing method based on black-level adjustment. The method includes steps of extracting luminance values of critical pixel points, carrying out sum calculation of luminance difference values of left and right channels and adjusting a reference level in an AD-chip analog-to-digital conversion process and the like so that it is ensured that luminances at the left side and at the right side of the spliced image are identical. The algorithm is executed every time when a frame of image is output by a CCD image sensor and adjustment is carried out in a real-time manner so that effects from scenario change can be deal with sufficiently. According to the method, the effect of a real scenario can be restored in a most truly manner after the images of the left and right channels are spliced.

Description

The image seamless joining method of adjusting based on black level

Technical field

The present invention relates to a kind of ccd image sensor image adjusting method of protection and monitor field high-definition camera, when the output image that relates in particular to a kind of binary channels ccd image sensor completes and splices after analog-to-digital conversion, the image seamless joining method of adjusting based on black level of removal of images YC deviation.

Background technology

CCD(Charge Coupled Device charge coupled device) imageing sensor is the core component of video camera.Current protection and monitor field, for meeting the requirements at the higher level to image definition, the pixel of ccd image sensor target surface is also more and more.But video frame rate can not reduce, so in the unit interval, ccd image sensor need to be exported more pixel, also just mean higher vision bandwidth.If output channel number is constant, transmission speed just need to increase considerably so, and the analog signal of high-speed transfer, it is pretty troublesome that integrality is processed, and therefore occurred binary channels ccd image sensor.

Binary channels ccd image sensor is divided into left and right two halves by image, by two passage outputting analog signals independently, through two independently AD chip respectively the analog signal of two passages is carried out to analog-to-digital conversion, quantification digital signal after conversion is sent to back-end processor, is completed the splicing work of left and right image by processor.

But in the circuit of practical application, because the electric property of two passages in left and right can not be in full accord, add the impact of AD chip periphery circuit.Therefore analog-to-digital result also has difference, and reaction, on image, is exactly left and right image deviation slightly in brightness.

Therefore need a kind of adjustment algorithm, reduce the impact of said external factor, thereby after assurance splicing, the left and right image of image is consistent in brightness.

In addition, in actual product, if this adjustment algorithm only comes into force when camera dispatches from the factory, and method of adjustment is solidificated in memory, when scene change and high low temperature change so, this adjustment is poor effect often, therefore need to often receive a two field picture, all the parameter of AD chip be adjusted.

Summary of the invention

The present invention is exactly in order to solve above-mentioned the problems of the prior art, and a kind of image seamless joining method of adjusting based on black level is provided.

In order to achieve the above object, the present invention adopts following technical scheme:

The image seamless joining method of adjusting based on black level of the present invention, comprises the steps:

(1) ccd image sensor gathers original image signal, and through left and right, two passages send respectively analog electrical signal to two AD chips;

(2) two AD chips carry out analog-to-digital conversion to the analog signal of receiving respectively, and digital signal is passed through to LVDS(Low Voltage Differential Signaling, low-voltage differential signal) interface sends FPGA(Field Programmable Gate Array, field programmable gate array to); FPGA is used two buffer areas to store respectively the digital signal from two AD chips;

(3) FPGA extracts respectively the rgb value of feature pixel from two buffer areas, by weighted average calculation, obtains the luminance difference of two passages in left and right;

(4) according to difference, fix the parameter of an AD chip, adjust another AD to input analog signal the reference level value while carrying out analog-to-digital conversion;

(5) adjust result came into force before next frame image arrives.

Advantage and good effect that the present invention has are:

The image seamless joining method of adjusting based on black level of the present invention, has eliminated the deviation of left and right image in brightness, the consistency of the right and left that has guaranteed image after final splicing in brightness.Every output one two field picture of ccd image sensor, all carries out once this algorithm, adjusts in real time, fully tackles the impact of scene change.According to said method, after the two channel image splicings of left and right, the effect of rediscover scene the most truly.

Accompanying drawing explanation

Fig. 1 is that hardware system of the present invention forms and signal flow graph;

Fig. 2 is the block plan of a two field picture of the present invention.

Embodiment

Below in conjunction with the drawings and specific embodiments, the image seamless joining method based on black level adjustment of the present invention is described further.Following each embodiment is not only limitation of the present invention for the present invention is described.

Fig. 1 has provided hardware system composition used in the present invention and signal flows to.Fig. 2 has described the subregion of a two field picture, and wherein, blank parts is effective district, and dash area is blanking zone, and middle solid line is the line of demarcation of CCD the right and left, and four dotted lines are the character pixel that extracts in the algorithm position in the effective district of image.

As shown in Figure 1, the image seamless joining method of adjusting based on black level of the present invention, comprises the steps:

(1) ccd image sensor gathers original image signal, and through left and right, 2. 1. two passages send analog electrical signal respectively to two AD chips.

(2) two AD chips carry out analog-to-digital conversion to the analog signal of receiving respectively, and by digital signal by 3. 4. two LVDS interfaces send FPGA to; FPGA is used two buffer areas to store respectively the digital signal from two AD chips.

(3) FPGA obtains feature pixel from two buffer areas, and feature pixel is respectively to extract two row pixels at CCD the right and left " seam crossing ", as shown in Fig. 2 dotted line position.Extract respectively the rgb value of feature pixel, by weighted average calculation, obtain the luminance difference of two passages in left and right.If effectively district has N capable, extract altogether 4N pixel.If the brightness value A of j the pixel that i is capable _ijrepresent, the relative brightness difference that i is capable is L _i=(A _i1+ A _i2)-(A _i3+ A _i4).L _ihigh for just representing left side brightness value, for negative indication the right brightness value is high, equal zero and represent that both sides brightness value equates.Luminance difference to all effective row is averaged, L=

.

(4) according to difference, fix the parameter of left passage AD, adjust the reference level value of right passage AD chip dies number conversion process.If L=0, does not make any adjustments.When L is greater than 0, the sampling reference level of right passage AD reduces by 1 unit, and when L is negative, the sampling reference level of right passage AD increases Yi Ge unit.Adjustment process is very fast, and concerning AD, valid data are sent, and when last two field picture enters blanking zone, changes AD parameter in blanking zone, can not affect the image of actual output.Before next frame image arrives, reference level value has been changed complete.

(5) adjust result and came into force before next frame image arrives, so far adjustment process finishes.By this step, progressively adjust, can make the image brightness of the right and left move closer to.As long as scene is constant, this process restrains, and image output tens frames per second, and for human eye, brightness regulation process is also very rapidly.

Claims (1)

1. an image seamless joining method of adjusting based on black level, is characterized in that, the method comprises the steps:

(1) ccd image sensor gathers original image signal, and through left and right, two passages send respectively analog electrical signal to two AD chips;

(2) two AD chips carry out analog-to-digital conversion to the analog signal of receiving respectively, and by LVDS interface, send digital signal to FPGA; FPGA is used two buffer areas to store respectively the digital signal from two AD chips;

(3) FPGA extracts respectively the rgb value of feature pixel from two buffer areas, by weighted average calculation, obtains the luminance difference of two passages in left and right;

(4) according to difference, fix the parameter of an AD chip, adjust another AD to input analog signal the reference level value while carrying out analog-to-digital conversion;

(5) adjust result came into force before next frame image arrives.

Images