Abstract: A method for detecting motion pixels in image includes the following steps. The difference image between current image and background image is firstly obtained. Then, a binarization threshold from the distribution of all absolute pixel values in the difference image is found. All or part of the pixel values, in difference image, whose absolute values are smaller than or equal to the binarization threshold is used to compensate the effect of illumination change and to obtain an updated difference image. The updated difference image is then binarized, and the binarized updated difference image whose values are high are determined motion pixels.

Abstract: Disclosed is a method and device for human face detection and recognition used in a preset environment. The human face detection device comprises a camera module, a synchronous image acquiring module, a human face detection module, a human face recognition module and a member face database module. The synchronous image acquiring module is used to acquire synchronously images photographed by the camera module. The human face detection module is used to detect if any human facial feature is presented in the image. If the human facial feature is confirmed, the image is transferred to the human face recognition module to extract the human feature from the image. Then, the extracted human facial feature is compared with member face data stored in the member face database module so that a recognition result which shows a person corresponding to the image is a legal or illegal member is generated.

Abstract: A string extraction method for an image based on multiple stroke width patterns matching is provided. The string extraction method includes matching each possible stroke width pattern with an input image for obtaining the similarities between the width patterns and each pixel, wherein the integrated similarity between each pixel and the stroke width patterns is namely an integrated computing result of all similarities between the pixel and all stroke width patterns.

Abstract: A method of facial features based human face recognition is disclosed. A human face and facial features thereof with respect to an input image corresponding to a person are first detected person by person by image processing technology. Then, each of such facial features for a plurality of persons are categorized into several categories and expressed to form a human facial features database for the plurality of persons. A to-be-searched or recognized human face image of a person is inputted. Then, the image is acquired with positions of the person's face and facial features by image processing technology and each of the facial features is categorized into several categories each with a specific expression. Then, according to the categories which the facial features of the person belongs to, the person may be recognized. As such, the purposes of human face search and recognition are achieved.

Abstract: An electronic ticket service system based on color scale image recognition for individual identification, purchasing certificate or bill payment. Different from the conventional black-and-white one- or two-dimensional barcode, the color scale uses assembly of color scales for various colors. Each assembly of multiple color scales represents one element, such that the information capacity is greatly increased. The user can order the ticket via phone or internet. When the user has successfully input the purchase information and complete the payment procedure, the electronic ticket service center can issue and transmit the electronic ticket containing the color scale code to the assigned device through multimedia or wireless transmission. The assigned device can be cellular phones or other mobile communication terminal devices. Upon entrance, the electronic ticket displayed by the cellular phones or the mobile communication terminals is verified by a ticket verification device.

Abstract: The image-based container defects detector consists of a plurality of camera units, a sensor unit, a frame grabber, and an image recognizer unit. The sensor unit serves to detect the vehicles entering the inspection station, and transmit a signal to the frame grabber for capturing the vehicle image by triggering the CCD camera thereby obtaining the information about the container's defect position and size. By dexterously utilizing image processing technique, the HIS (hue, saturation and intensity) color is employed to distinguish a normal area from a defected area in the container. Then quad-tree and merge is used to segregate image roughly. As for the non-defected area becoming obscured due to noise is removed using a filter. Finally the defected area is displayed on the screen. The present invention is well-suited for after detection as well as for on-line immediate detection for defect containers.

Abstract: A remote video-on-demand digital monitoring system principally includes an image capture module, a server relay module and a video-on-demand module. The image data of said image capture module are stored in the server relay module through a telecommunication network. Through the telecommunication network, customers can browse the monitored web pages of the video-on-demand module via a browser, and view or store the monitored image data at a desired time segment according to a video-on-demand or movies-on-demand mode.

Abstract: A method for detecting motion pixels in image includes the following steps. The difference image between current image and background image is firstly obtained. Then, a binarization threshold from the distribution of all absolute pixel values in the difference image is found. All or part of the pixel values, in difference image, whose absolute values are smaller than or equal to the binarization threshold is used to compensate the effect of illumination change and to obtain an updated difference image. The updated difference image is then binarized, and the binarized updated difference image whose values are high are determined motion pixels.

Abstract: An automatic car toll paying method is provided. Before a car enters a toll road or bridge, the user previously calls a specific number representing a log in of an automatic charging through a on-vehicle cellular communication equipment and then hangs up, and after leaving the toll road, the user calls a specific number representing a log out of the automatic charging. Continuously, wherein said on-vehicle cellular communication equipment has been logged in automatic charging when said car enters cell governing area of a communication coverage covering a toll station, the cellular base station will inform the billing center to automatically chargeback from an account connected to the on-vehicle cellular communication equipment or record the toll of the passing car.

Abstract: A string extraction method for an image based on multiple stroke width patterns matching is provided. The string extraction method includes matching each possible stroke width pattern with an input image for obtaining the similarities between the width patterns and each pixel, wherein the integrated similarity between each pixel and the stroke width patterns is namely an integrated computing result of all similarities between the pixel and all stroke width patterns.

Abstract: An innovative method of automatic vehicle licence plate recognition is described, the vehicle licence plate recognition procedure is carried out respectively with reference to three component color red(R), green(G), and blue(B) grey level pixel maps, and then the results are reorganized. Sine the synthesized three component color R, G, B grey level pixel map is obtained by mixing R, G, B color component according to a definite proportion (for example dividing the sum of three component by 3), hence among the three component color grey level pixel maps, there must be at least one (R, or G, or B) whose contrast is better than that of the synthesized one. This most confidential one is picked out as the final result of the licence plate recognition.

Abstract: An innovative method of automatic container number recognition is described, the container number recognition work is carried out respectively with reference to three component color red(R), green(G), and blue(B) grey level pixel maps, and then the identified results are reorganized. Since the synthesized three component color R, G, B grey level pixel map is obtained by mixing R, G, B color component according to a definite proportion (for example dividing the sum of three components by 3), hence among the three component color grey level pixel maps, there must be at least one (R, or G, or B) whose contrast is better than that of the synthesized one. This most confidential one is picked out as the final result of the container number recognition.

Abstract: The image-based container defects detector consists of a plurality of camera units, a sensor unit, a frame grabber, and an image recognizer unit. The sensor unit serves to detect the vehicles entering the inspection station, and transmit a signal to the frame grabber for capturing the vehicle image by triggering the CCD camera thereby obtaining the information about the container's defect position and size. By dexterously utilizing image processing technique, the HIS (hue, saturation and intensity) color is employed to distinguish a normal area from a defected area in the container. Then quad-tree and merge is used to segregate image roughly. As for the non-defected area becoming obscured due to noise is removed using a filter. Finally the defected area is displayed on the screen. The present invention is well-suited for after detection as well as for on-line immediate detection for defect containers.

Abstract: An automatic car toll computing and charging method is provided. The method includes steps of employing a cellular phone to process a log in and a log out of a car toll computing; previously storing the connection between a cellular phone number and a vehicle identification number, or keying in said vehicle identification number or other car IDs sufficient to bring out said vehicle identification number in a billing center; and calling a specific number representing a log in of an automatic computing and charging through an on-vehicle cellular communication equipment and before a car enters a toll road or bridge and hanging up, and calling a specific number representing a log out of the automatic computing and charging after leaving the toll road or bridge. Consequently, the toll can be charged through calculating all cell IDs which by a cellular phone system passes theretrough during the log in and the log out.

Abstract: A personal identification system based on the reading of multiple one-dimensional barcodes scanned from PDA/cell phone screen, which is a personal identification system applicable for reading multiple one-dimensional barcode on PDA/cell phone screen, comprises a multiple one-dimensional barcode generating system and a reading system of multiple one-dimensional barcode on PDA/cell phone screen, wherein said multiple one-dimensional barcode generating system is consisted of a data input unit and a barcode generating unit. The reading system of multiple one-dimensional barcode on PDA/cell phone screen is consisted of a PDA/cell phone reaction unit, a CCD camera, a frame grabber, an image processing unit, a barcode recognition unit and a data recording unit. The personal identification system based on the reading of multiple one-dimensional barcodes scanned from PDA/cell phone screen is useful as the mechanism for personal identification, certificates of purchase or checkoff.

Abstract: A digital image monitoring system with functions of motion detection and auto iris which aims at the requirement of video-recording monitoring for traffic facilities, home or company which is composed of CCD camera, lens, frame grabber, illumination—reflectance decomposition unit, illumination variation detection unit and motion detection unit. The frame grabber accompanies with the CCD camera and lens to take digital images continuously. Then the illumination—reflectance decomposition unit decomposes the captured digital image and transfers it into the illumination and reflectance spectrum to provide the illumination detection unit to analyze the illumination spectrum variation between two adjacent image frames and thus this spectrum variation is used to control the amplification or reduction of the iris; beside ,an object movement detection unit is provided to analyze the illumination variation of two adjacent films to judge if the object in the frame is moving .

Abstract: Data samples describing a plurality of micro-segments that compose a symbol to be recognized are received from a device such as an electronic pad. A preprocessor maps the micro-segments into cells of an array that has several feature dimensions. The preprocessor assigns values to the cells based on the length of a micro-segment associated with the cell, and how well the features of the associated micro-segment correspond to the feature label of the cell. The cell values are used as inputs to at least one of a plurality of neural networks, where each neural network is trained to identify symbols from a different group of symbols. A sorter examines the symbol to determine which neural network should be used to recognize the symbol. An output produced by the sorter controls a switching means that communicates the cell values to the proper neural network.

Abstract: Data samples describing a plurality of micro-segments that compose a symbol to be recognized are received from a device such as an electronic pad. Preprocessors map the micro-segments into cells of a plurality of feature arrays with different resolutions. Preprocessors assign values to the cells based on the length of a micro-segment associated with the cell, and how well the features of the associated micro-segment correspond to the feature label of the cell. The cell values are used as inputs to comparators that compare the feature arrays with reference arrays. The results of comparisons involving lower resolution feature arrays and reference arrays, are used to limit the number of comparisons involving higher resolution feature arrays and reference arrays. The highest resolution comparison selects the reference array that identifies the symbol to be recognized.