CN102622726A

CN102622726A - Multiple-display-mode-supporting gray scale conversion image enhancement processing

Info

Publication number: CN102622726A
Application number: CN2011100299698A
Authority: CN
Inventors: 周均杰; 雷西军; 陈�峰
Original assignee: CHANGZHOU XINQI MICROELECTRONICS TECHNOLOGY Co Ltd
Current assignee: CHANGZHOU XINQI MICROELECTRONICS TECHNOLOGY Co Ltd
Priority date: 2011-01-27
Filing date: 2011-01-27
Publication date: 2012-08-01

Abstract

The invention belongs to the field of digital image processing. According to the invention, gray scale conversion is specifically carried out on the dark part of the digital image, that is, the gray value of a part which has a small gray value in the image is increased through gray mapping, and meanwhile, the gray value of a highlight area is maintained, so that an image enhancement on the dark area of the image is realized and the dark area which is easy to be overlooked and hard to be identified becomes easy to be observed. The gray scale conversion provided by the invention can be applied to multiple display modes, such as SCTV, HDTV, LCD, and the like, and have a high flexibility and compatibility.

Description

Support the gradation conversion image enhancement processing of many display modes

One. technical field

Patent of the present invention belongs to Digital Image Processing, Flame Image Process integrated circuit, Digital Image Processing algorithm field.Handle to coloured image in the digital picture and gray level image, be used to strengthen the visual effect of image, make some originally uncared-for image detail arrived by people are observed more easily, digital picture analyzed and watched thereby be convenient to people.This patent is not limited to a certain image display pattern, and it has higher flexibility and compatibility applicable to multiple demonstrations such as SDTV, HDTV, PC, LCD.

Two. background technology

Continuous development along with IC industry; On a large scale; The manufacturing of VLSI (very large scale integrated circuits) is increasingly mature; Its price was but in continuous reduction when chip functions became from strength to strength, thereby made all kinds of consumption electronic products such as DC (digital camera), DV (DV), PC (PC) universal day by day, and people are also improving constantly for the quality requirements of digital picture.

Many times, the digital picture that people obtain always exists variety of issue, and such as owing to reasons such as illumination, exposures, digital picture is whole dark partially, and some details are left in the basket under the situation that people extremely do not examine to such an extent as to image is crossed secretly; Digital picture integral body is bright partially, to such an extent as to thereby some place lets the people feel that the dazzling people of influence are to image observation.Therefore seek a kind of simple effective method improve digital picture quality become very urgent be necessary.

Three. summary of the invention

This patent is mainly used in the enhancing of dark areas digital picture.Because the gray-level difference that can differentiate of human eye is about 1%, when we adopted 8 bit data to represent a chrominance component, the differentiable gray-level difference of human eye was about 3 grades, but in reality, human eye for cross dark areas and the separating capacity of crossing bright area a little less than.Usually can think that 10 grades of following human eyes of gray scale can't differentiate, in view of this consideration, under the situation that as far as possible keeps contrast, in this patent, 0 grade of gray scale still is mapped as 0 grade of gray scale, be 10 grades of gray scales with 3 grades of grey scale mapping.Between 10 grades of gray scale to 20 grade gray scales; It is two gray levels of 5 that human eye can be distinguished gray scale difference; Therefore this patent is 6 grades of grey scale mapping 15 grades of gray scales, is 9 grades of grey scale mapping 20 grades of gray scales, and between 20 grades of gray scale to 30 grade gray scales human eye can to differentiate gray scale difference be two gray levels of 4; Therefore this patent is 12 grades of grey scale mapping 24 grades of gray scales; 15 grades of grey scale mapping are 28 grades of gray scales, and can distinguish gray-level difference at 30 grades of above human eyes of gray scale is two gray levels of 3, so this patent is 18 grades of grey scale mapping 32 grades of gray scales.And when gray level was big, digital picture had met the eye-observation demand basically, therefore its gray level was no longer adjusted.This patent adopts the YCbCr color space that digital picture is adjusted.Because general pattern transmits with the RGB color space, so this patent adopts following transformational relation:

Y＝0.183R+0.614G+0.062B+16

Cb＝-0.101R-0.338G+0.439B+128

Cr＝0.439R-0.399G-0.040B+128

R＝1.164(Y-16)+1.793(Cr-128)

G＝1.164(Y-16)-0.534(Cr-128)-0.213(Cb-128)

B＝1.164(Y-16)+2.115(Cb-128)

Four. description of drawings

Fig. 1 is the function schematic block diagram of gradation conversion, and wherein y_in, cb_in, cr_in, de_in, vsync_in, hsync_in are input signal, and y_out, cb_ou, cr_out, de_out, vsync_out, hsync_out are the output signal.

Fig. 2 is the sequential chart of gradation conversion, owing to adopt mapping method, the output signal is a clk with respect to the time delay of input signal.。

Fig. 3 is that this patent proposes first kind the input-output curve that is used for gradation conversion.

Fig. 4 is second kind of input-output curve that is used for gradation conversion that this patent proposes.

Five. embodiment

This patent has a control signal and is used to control the color that whether carries out image and strengthens, and when control signal is a high level, control signal is effective, and the color that carries out image strengthens.Through the Y value of input data is judged, seek the output data of corresponding input Y.In practical application, we have chosen two groups of gray scale adjustment curves Y-signal is carried out the gray scale adjusting: first group is mapped as 17,17 with 16 and is mapped as 21; 18 are mapped as 23,19 is mapped as 27,20 and is mapped as 29; 21 are mapped as 30; Adopt curve Y=(6910-37*X) * X/4200 to obtain the grey scale mapping of Y from 22 to 31, adopt curve Y=1.1329*X-0.0029*X*X+11.8226 to obtain the grey scale mapping of Y from 32 to 90, later grey scale mapping is itself in Y＞91; Second group of grey scale mapping that adopts curve Y=(6910-37*X) * X/4200 to obtain Y from 0 to 70; Adopt curve Y=1.0756*X-4*X*X/6750 to obtain the grey scale mapping of Y from 71 to 90; Adopt curve Y=0.9*X+11 to obtain the grey scale mapping of Y from 91 to 110, it is mapped as itself when the later gray level in Y＞110.

The implementation of this patent can be to realize as adopt FPGA or make ASIC through hardware, also can realize as write c program through software approach, no matter adopt above which kind of mode all to belong to scope.

This patent adopts direct mapping method to realize input and directly conversion of output; It is the output Y value that the Y component of each input can search out a correspondence through the mode of look-up table; But because input can obtain through curve with the direct relation of output; Obtain output so when concrete the realization, can not adopt direct mapping through the curve evaluation, this belongs to scope equally.The relation curve of input and output obviously is not unique in addition, thus for the adjusting trend of high eight-bit data or for the incremental adjustments trend of Y identical with this patent all belong to scope.Suspension points is represented the increase successively of data.

8	13	8
			9	14	9
10	16	10
			11	17	11
12	18	12
			13	20	13
14	21	14
			15	23	15
17	24	16
			21	25	17
23	27	18
			27	28	19
29	29	20
			30	31	21
32	32	22
			33	33	23
34	34	24
			35	36	25
37	37	26
			38	38	27
39	39	28
			40	40	29
42	41	30
			43	43	31
45	44	32
			46	45	33
47	46	34
			48	47	35
49	48	36
			50	49	37
51	50	38
			52	51	39
52	52	40
			53	53	41
54	54	42
			55	54	43
56	55	44
			57	56	45
58	57	46
			59	58	47
60	59	48
			60	59	49
61	60	50
			62	61	51
63	62	52
			64	62	53
65	63	54
			65	64	55
66	65	56

67	65	57
			68	66	58
69	66	59
			69	67	60
70	68	61
			71	68	62
72	69	63
			72	69	64
73	70	65
			74	70	66
75	71	67
			75	71	68
76	72	69
			77	72	70
78	73	71
			78	74	72
79	75	73
			80	76	74
80	77	75
			81	78	76
82	79	77
			83	80	78
83	81	79
			84	82	80
85	83	81
			85	84	82
86	85	83
			87	86	84
87	87	85
			88	88	86
88	89	87
			89	90	88
90	91	89
			90	92	90
91	93	91
			92	94	92
93	95	93
			94	96	94
95	97	95
			96	97	96
97	98	97
			98	99	98
99	100	99
			100	101	100
101	102	101
			102	103	102
103	104	103
			104	105	104
105	106	105

106	106	106
			107	107	107
108	108	108
			109	109	109
110	110	110
			…	…	…
255	255	255

Claims

1. given input data and output data relation is sought the output Y value of correspondence according to input pixel Y value, keeps the constant image enhancement processing method that is primarily aimed at the digital pattern dark areas of input and output Cb and Cr value simultaneously.

2. two groups are used for the data that gray scale is regulated.

3. the method for article one description comprises the ASIC of hardware handles aspect, the Computer Image Processing software of software processes aspect in the application of different field.

4. the application of method aspect the difference demonstration of article one description comprises SDTV, HDTV, LCD.

5. the input/output relation of second description comprises the ASIC of hardware handles aspect, the Computer Image Processing software of software processes aspect in the application of different field.

6. the application of input/output relation aspect the difference demonstration of second description comprises SDTV, HDTV, LCD.