Grayscale

About: Grayscale is a research topic. Over the lifetime, 13278 publications have been published within this topic receiving 156084 citations. The topic is also known as: grayscale image & black-and-white image.

Device for processing image data

Abstract: PURPOSE: A device for processing image data is provided to process image data, by adjusting an image revival device, which displays a pixel area to a gradation level of over two steps. The pixel area raises a gradation revival ability for an image of the image data expressed as a gray scale, and makes a resolution not greatly lower. CONSTITUTION: A device for processing image data comprises a multi-level halftone processor(20), a serial/parallel converter(30), and a digital pulse modulator/demodulator(40). The multi-level halftone processor(20) converts an image data inputting gray scale information for a pixel into each pixel gradation signal to process, through a converting method of predetermined data, by being corresponded to gradation step numbers expressing a pixel area by an image revival device. The serial/parallel converter(30) outputs each pixel gradation signal in parallel as many as predetermined pixels. Each pixel gradation signal corresponds predetermined pixel numbers successively output serially in the multi-level halftone processor. And the digital pulse modulator/demodulator(40) outputs a pulse modulating signal corresponding to a chosen pulse modulating mode to the image revival device, by inputting a gradation level value of gradation signals to an input value.

Predistortion head up display image anti-aliasing method based on edge direction correlation

Abstract: The invention relates to a predistortion head up display image anti-aliasing method based on edge direction correlation and belongs to the field of digital image processing. The method includes usingFIFO for row buffering and column buffering of an input DVI image, and extracting 9 pixel values in a 3X3 window; using a quick sorting method to find a maximum value and a minimum value for 9-point pixel data in a template window; using maximum and minimum values in a 3X3 template to determine the correlation of data in a template, and distinguishing an edge region and a non-edge region of the image; for the edge region of the image, using a direction template to determine the edge direction and then performing the weighting process after the distance parameter normalization; for the non-edgeregion, performing direct delay output, and causing the delay time to be consistent with the anti-aliasing processing time; and finally performing gray-scale transformation on overall character images by utilizing a segmentation function curve according to inherent characteristics of a head up display raster image. According to the method, the display quality of the head up display raster image can be improved as a whole, and the gray scale loss and the brightness loss caused by the anti-aliasing process are reduced.

NETD automatic testing method for infrared thermal imager

Abstract: The invention provides an NETD automatic testing method for an infrared thermal imager. The method comprises the steps of (1) acquiring data of a frame of target image, and performing Gaussian filtering denoising smoothing processing on the acquired image data; (2) converting a denoised grayscale image into a binary image by using a threshold transformation method; (3) calculating out mass centercoordinates of all connected domains of the binary image; and (4) finding out the connected domain with the minimum difference value between a long axis and a short axis in the connected domains, wherein the connected domain is an area where a target is located; and then calculating out a noise value and a signal transmission function of the target area according to the position and the size of the area where the target is located, wherein the ratio of the noise value to the signal transmission function is the NETD value of the infrared thermal imager. According to the method, the size of thetarget can be automatically recognized by utilizing the image processing method, and meanwhile, the size of the target is calculated out, so that it is not required to artificially observe the effective area of the target and manually select area data for performing NETD correlation calculation.

Method and apparatus for recognizing objects in images broadcast, their location and reading their postal details

Abstract: It performs the following steps: - converting in the respective color camera (11) color image formed in a grayscale image with respect to the color camera (11) higher, to read the postal data in the automatic reading devices suitable resolution, - converting in the respective color camera (11) color image formed in a color image with respect to the camera resolution lower, for recognizing objects and their position sufficient resolution, - detecting and identifying the objects in a respective color lower resolution image, including its location, wherein the assignment of the objects are postal to specified categories based on established in a preceding instruction phase structure rules, - Automatic reading of the required postal information in the gray scale images of the relevant, identified objects.

Time domain periodic point target detection method

Abstract: The invention discloses a time domain periodic point target detection method comprising the following steps that the single frame original image of an object to be detected is converted into a grayscale image, and noise suppression is performed on the grayscale image so as to obtain the denoised image; double differential joint binarization processing is performed on the denoised image so as to obtain a candidate target binary image; connected domain marking is performed on the candidate target binary image so as to obtain a single frame candidate target point set; spatial domain feature analysis and extraction are performed and the false target points of the single frame candidate target point set are eliminated; time domain feature analysis and extraction are performed so as to obtain ahighly suspected target point set; and the highly suspected target of the longest duration acts as the real target through multi-frame information comprehensive judgment The false alarm probability of single frame detection can be reduced, the calculation complexity of the multi-frame detection algorithm subsequently using the time domain features can be effectively reduced and the false target points which meet the spatial domain features and are difficult to be eliminated by single frame detection can be effectively filtered and thus point target detection can be realized