Summary of the invention
In order to solve the problems referred to above that prior art exists, the invention provides a kind of method that infrared medicine image and sliced image of human body merge.
The technical solution adopted in the present invention is: a kind of method that infrared medicine image and sliced image of human body merge, and it comprises the following steps: obtain human body infrared sequence image and CT sequence image; Human body infrared image obtains area-of-interest; Human body CT sequence image is split, obtains 3 D human body sectioning image; According to the area-of-interest on human body infrared image, human body infrared image and sectioning image are carried out coordination according to a certain percentage corresponding; According to 3 D human body sectioning image adjustment area of space, the human body infrared image and sectioning image coordinating correspondence is cut; Regulate the transparency of the human body infrared image after cutting and sectioning image, adopt Image Fusion that the human body infrared image after adjustment and sectioning image are carried out multi-modal fusion, obtain fused images.
Further, Image Fusion is adopted to carry out multi-modal fusion to the human body infrared image after adjustment and sectioning image, obtain fused images, its process is: according to image pixel, and human body infrared image and sectioning image are carried out additive fusion with adjustable weight proportion.
Further, Image Fusion adopts Wavelet Fusion method or laplacian pyramid fusion method.
Further, employing light projecting algorithm to the implementation procedure that CT sequence image carries out volume drawing is: play up forward face depth map, obtain forward depth texture; Face depth map is dorsad played up, obtains depth texture dorsad; Vertex position and directions of rays is calculated in vertex shading program; In fragment shader program, retrieve forward depth texture, obtain closely; In fragment shader program, retrieve depth texture dorsad, obtain remote; Calculate the maximum traverse distance on current ray direction; In fragment shader program, calculate sample texture coordinate, and volume textures is sampled; In fragment shader program, carry out color synthesis and transparency cumulative; In fragment shader program, judge whether sampled distance exceedes ultimate range or transparency is cumulative whether more than 1; Then in fragment shader program, recalculate sample texture coordinate if sampled distance does not exceed ultimate range or transparency is cumulative more than 1, and again volume textures is sampled, and order performs; Otherwise, export color value.
Further, the method that described infrared medicine image and sliced image of human body merge is further comprising the steps of: by carrying out D translation to the human body infrared image after cutting and sectioning image, show the internal organs in corresponding human body three-dimensional space.
Further, the method that described infrared medicine image and sliced image of human body merge is further comprising the steps of: place radioactive source, the fusion image data obtained and radioactive source is carried out fusion and draws, and preserves merging the view data of drawing.
Owing to adopting above technical scheme, beneficial effect of the present invention is: the view data of the same target that dissimilar sensor can obtain by the inventive method carries out spatial registration, and adopt certain Image Fusion, Information Superiority contained by each view data or complementarity are organically combined, produces new view data; Thus be convenient to the human dissection position of determining that infrared image is corresponding, be convenient to the relation determined between human surface temperature and human internal organs, make infrared image medically can quantize application and judge.
Embodiment
Below in conjunction with drawings and Examples, the present invention is described in detail.
As shown in Figure 1, the invention provides a kind of method that infrared medicine image and sliced image of human body merge, it comprises the following steps:
1) human body infrared sequence image and CT sequence image is obtained.
2) on human body infrared image, obtain ROI (RegionOfInterest, area-of-interest) region.
3) human body CT sequence image is split, obtain 3 D human body sectioning image.
4) according to the area-of-interest on human body infrared image, human body infrared image and sectioning image are carried out coordination according to a certain percentage corresponding.
5) according to 3 D human body sectioning image adjustment area of space, the human body infrared image and sectioning image coordinating correspondence is cut.
6) transparency of the human body infrared image after cutting and sectioning image is regulated, Image Fusion is adopted to carry out multi-modal fusion to the human body infrared image after adjustment and sectioning image, obtain fused images, its process is: according to image pixel, and human body infrared image and sectioning image are carried out additive fusion with adjustable weight proportion.Wherein, Image Fusion adopts Wavelet Fusion method or laplacian pyramid fusion method.
Above-mentioned steps 3) in, as shown in Figure 2, employing light projecting algorithm to the implementation procedure that CT sequence image carries out volume drawing is:
Forward face depth map is played up, obtains forward depth texture; Face depth map is dorsad played up, obtains depth texture dorsad.
Vertex position and directions of rays is calculated in vertex shading program.
In fragment shader program, retrieve forward depth texture, obtain closely.
In fragment shader program, retrieve depth texture dorsad, obtain remote.
Calculate the maximum traverse distance on current ray direction.
In fragment shader program, calculate sample texture coordinate, and volume textures is sampled.
In fragment shader program, carry out color synthesis and transparency cumulative.
In fragment shader program, judge whether sampled distance exceedes ultimate range or transparency is cumulative whether more than 1.Then in fragment shader program, recalculate sample texture coordinate if sampled distance does not exceed ultimate range or transparency is cumulative more than 1, and again volume textures is sampled, and order performs.Otherwise, export color value.
In above-described embodiment, the method that infrared medicine image of the present invention and sliced image of human body merge is further comprising the steps of: by carrying out D translation to the human body infrared image after cutting and sectioning image, show the internal organs in corresponding human body three-dimensional space.In addition, the health anomalies point of certain depth and size can also be set on human body infrared image, at three dimensions, health anomalies point be shown.
In above-described embodiment, in order to obtain more accurately, fused images clearly, the method that infrared medicine image of the present invention and sliced image of human body merge is further comprising the steps of: place radioactive source, the fusion image data obtained and radioactive source being carried out fusion draw, preserving merging the view data of drawing.
As shown in Figure 3, volume data and solid carry out fusion display, and its basic implementation procedure is:
1) be loaded into solid, generate volume textures carrier.
2) geometry data is plotted in FBO (frame buffer object FrameBufferObject) and preserves.
3) be loaded into volume data, obtain the bounding box of volume data, so as to carry out volume data cut drafting.
4) carry out respectively forward face depth map play up and dorsad face depth map play up, by the forward depth map of acquisition and dorsad depth map be all saved in corresponding FBO.
5) according to solid and step 4) the forward depth map that obtains and depth map dorsad, redefine entry position and the injection position of volume drawing ray, be kept in corresponding color and vein and depth texture.
6) according to step 5) in obtain new to inject, penetrate the volume drawing that volume data is carried out in position, be saved in corresponding FBO.
7) by step 2) and step 6) in obtain FBO carry out fusions drafting.
8) output display is carried out by finally merging the FBO completed.
The view data of the same target that dissimilar sensor can obtain by image interfusion method of the present invention carries out spatial registration, and adopt certain Image Fusion, Information Superiority contained by each view data or complementarity are organically combined, produces new view data.Compared with single piece of information source, this new view data has the information representation comparatively optimized describing institute's research object, it can reduce or suppress polysemy, imperfection, uncertainty and the error that may exist in perceived object or environment interpretation, maximally utilises the information that various information source provides.Image co-registration is divided into Pixel-level, feature level, decision level three levels, that wherein pixel-level image fusion can provide the fusion treatment on other level not have is abundanter, more accurately, more reliable detailed information, be conducive to the further analysis of image, process and understanding.
The present invention is not limited to above-mentioned preferred forms; anyone can draw other various forms of products under enlightenment of the present invention; no matter but any change is done in its shape or structure; every have identical with the application or akin technical scheme, all drops within protection scope of the present invention.