33results about How to "In line with observation habits" patented technology

The invention belongs to the field of computer-generated holography and three-dimensional display, and particularly relates to a time-division multiplexing-based color holographic three-dimensional display method and system. By the display system, a multi-view two-dimensional image of a color target object is obtained by a method of CCD camera or computer graphics; a group of target images suitable for phase-type hologram calculation are generated after a preprocessing transformation process; a corresponding phase-type hologram is calculated by an iterative Fourier transform algorithm and is loaded to a spatial light modulator in a time-division manner; a synchronization control circuit controls red, green and blue three-color light sources to synchronously display in sequence; and three monochromatic holographic reconstructed images are generated on a reconstruction plane in a time-division manner, and corresponding reconstructed image pixels are directionally diffracted to the position at which a pixel-type nano grating is fixed to form different viewpoints, thereby obtaining multi-view color reconstructed images and achieving true color three-dimensional display. The hologram is quickly refreshed and loaded by combining the spatial light modulator, so that the multi-view real color full-parallax dynamic holographic three-dimensional reconstruction of the target object is achieved.

The invention provides an interactive image processing method used for a video. The method comprises the following steps of extracting key frames from a video sequence; overlaying frames between adjacent key frames one by one to form a key frame cluster used for interactive marking; distinguishing the images of the key frames into a foreground area, a background area and an unknown area; carrying out spectral clustering and alpha value estimation on the key frames to obtain the sectional drawing result of the key frames; and finally, transferring the sectional drawing result of the key frames to the whole video sequence to obtain a final sectional drawing result. Because the foreground points are interactively marked on each key frame cluster, spatiotemporal information contained in the video sequence is fully utilized, and the naturalness and the intuition of user interaction are guaranteed on the basis of the user interface of the key frame, thereby conforming to the observation habit of a human visual system. Marking on the key frame clusters replaces independently marking on the key frames, and therefore the method has strong robustness for the situation with the big foreground object local action.

The invention belongs to the technical field of technical vision, and particularly relates to a method for steganography and reading of information in an image. According to the method, information ishidden in a non-salient region and a high-frequency region of an image to improve the human perception quality of a coded image; and meanwhile, the detector and the encoding/decoding system are trained together, so that the decoding system can automatically adapt to the positioning error of the detector to replace manual selection of the analog noise intensity, and finally, the balance between the robustness of the decoding system and the image quality is achieved.

The invention discloses a polarization colorization method based on deep learning. The polarization colorization method comprises the following steps: collecting images in four polarization directionsand color images corresponding to the same scene by adopting a polarization imaging system; calculating a polarization angle image and a polarization degree image of the acquired image, and calculating a polarization factor feature image corresponding to the scene; constructing a training set according to the color image of the scene and the corresponding polarization factor feature image; constructing a dual-flow architecture network of low-light polarization colorization based on deep learning; designing a loss function; and inputting the images in the training set into a deep neural network for training to obtain a polarization colorization model. According to the method, a colorized network with a double-flow architecture is adopted, global information and local information are fused,and sufficient extraction of image features is realized; human interference is avoided, the colorization effect of polarization imaging can be achieved, the observation habit of human eyes is met, and detail information of the scene is further enriched.

The present invention provides a head-mounted display device comprising a display screen which is a monochromatic or multichromatic organic light-emitting display screen having good light transmittance and through which a real scene can be seen in a non-lit or lit state, and an image conversion optical device that is used for converting an image generated by the display screen to the human eye visible distance and comprises one or more layers of microlens arrays; and one microlens of each microlens array corresponds to one or more pixels in the display screen. The image production system is avirtual-real fusion device, and the light weight of a head-mounted display scheme can be realized; in the 3D display scheme of the device, the nanostructures of the left and right light field lenses are symmetrical in a distributed manner, and a virtual scene generated between the left and right light field lenses has parallax, so that a binocular parallax effect is formed, and the observation habit of human eyes is met; the device has the advantages of few device units, simple structure and low manufacturing cost.

The invention discloses a stereoscopic-image visual significance extraction method based on frequency domain sparse representation. The method is characterized by acquiring the small-sized three-channel image of the left-viewpoint image of a testing stereoscopic image and a small-sized normalized parallax image; carrying out partitioning on the three-channel image and the normalized parallax imageand acquiring matrixes corresponding to partitioning images; using the four matrixes to form a quaternion matrix, acquiring a transformed quaternion matrix after two-dimensional quaternion Fourier transform, and according to the low frequency components of a real portion and three imaginary portions, acquiring a low frequency component image; using a sparse expression dictionary to extract the sparse weight matrix of the low frequency component image; according to the sparse weight matrix and a center preference image, acquiring two center and periphery significance images; carrying out fuzzyprocessing on the two center and periphery significance images and then fusing and acquiring a fusion image; using the center preference image to reinforce the center and periphery of the fusion image; and acquiring a visual significance image after center and periphery reinforcement image size conversion. The method has advantages of high extraction stability and good extraction accuracy.

The invention belongs to the technical field of line patrol robots and discloses a first perspective line patrol robot, comprising a line patrol robot platform, a pan-head system, a binocular camera and a pair of VR glasses, wherein the pan-head system is arranged on the line patrol robot platform and receives the head-rotation control information sent by the VR glasses, and is provided with a movable mechanism driven by a motor, which is used for controlling the tilt and left-right rotation of the mounted binocular camera to realize the first perspective shooting; the VR glasses can obtain the head rotation control information of an operator and send the head rotation control information to the pan-head system; and the VR glasses operator controls the pan-head movable mechanism to pitch upward and downward, left and right, and rotates the corresponding angle left and right. The first perspective line patrol robot solves the problems that the camera cannot follow the head of the operator to rotate during line patrol and the visual angle conversion process is complex, realizes the line patrol of the first visual angle, solves the problem that the photographed image has no stereoscopic feeling, enables the operator to find out the line fault more easily, and improves the patrol efficiency.

The invention provides a bidirectional histogram equalization processing method and system for infrared images. The input infrared images are processed to form histograms, high and low boundaries are further calculated, image gray center-of-gravity positions are calculated by the aid of a preset platform value and the high and low boundaries, and then shades and highlights divided by the center-of-gravity positions are equalized respectively, thus, the calculation method is higher in adaptability and has obvious advantages in processing images with large backgrounds and small target details, the data calculation amount and the design cost are effectively decreased, and meanwhile, through analysis of centers of gravity of the images, and highlight target details of the infrared images are embodied more obviously and relatively conform to the human eye observation habit; besides, compared with existing platform histogram equalization processing methods and systems, the method and the system have the advantages that the processed highlight target details are not excessively exaggerated, and embodiment of other details of the shades of the infrared images is facilitated.

The invention provides an image processing method and device, and the method comprises the steps: photographing a first image through a camera module under the condition that the camera module is separated from an equipment body, and obtaining first posture information; upon receiving a correction instruction, displaying the second image through the electronic equipment, wherein the second image is obtained by performing correction processing on the first image according to the first attitude information and second attitude information of the electronic equipment, wherein the correction processing comprises the steps of determining a target rotation angle along a target rotation direction according to the first attitude information and the second attitude information, and rotating the first image along the target rotation direction at the target rotation angle to obtain the second image. According to the invention, the first image can be rotated along the target rotation direction at the target rotation angle to obtain the second image for display, so that the image actually viewed by the user on the electronic equipment conforms to the observation habit of the user, and the display effect of the image is improved.

The invention discloses intelligent blind-assisting wearable equipment controlled by a voice assistant. The intelligent blind-assisting wearable equipment comprises glasses and a wristband; an ultrasonic ranging module is installed in the middle of the front side of a glasses frame, cameras comprising AI object searching modules and visual recognition modules are arranged in the left side and theright side of the glasses frame respectively, a first Bluetooth module, a loudspeaker, a picture processing chip and a main control chip are installed on the left glasses leg and the right glasses legrespectively, the two ends of a glasses lantern ring are fixedly connected with the left glasses leg and the right glasses leg respectively, and the lantern ring is provided with a microphone, a volume key for adjusting the volume of the microphone, a first storage battery, a first startup and shutdown key and a first charging hole; the wristband comprises a main body and a fixing sleeve ring, aBeidou positioning module, a second Bluetooth module and a second storage battery are installed in the main body, an SIM card slot, a second startup and shutdown key and a second charging hole are formed in the side face of the main body, and the second Bluetooth module is in wireless communication with the first Bluetooth module. The intelligent blind-assisting wearable equipment controlled by the voice assistant helps the blind to go out, study and live, and the life of the blind is not restrained any more.

The invention discloses an adjusting screw mechanism. The mechanism is characterized in that a gasket is fixedly arranged on an adjusting screw, a flat tail of the adjusting screw is arranged in a square hole of an inner adjusting handle, a pin is arranged in the inner adjusting handle and is matched into a waist-shaped hole of the adjusting screw, and the pin is fixed to the inner adjusting handle by tightening a screw. A positioning spring, a positioning pin and a steel ball are sequentially arranged in a blind hole of the inner adjusting handle, the spring tightly presses the positioning pin, the steel ball placed in a positioning pin hole makes contact with a triangular inner tooth bevel groove of a hand wheel base, external threads of the adjusting screw for assembly are in threaded connection with internal threads of the hand wheel base, external threads of a hand wheel check ring are in threaded connection with the internal threads of the hand wheel base, the inner adjusting handle is limited, and the hand wheel check ring is fixed to the hand wheel base by tightening a screw, an outer adjusting handle is fixed to the inner adjusting handle by tightening a screw, an inner sealing ring is arranged between the hand wheel base and the inner adjusting handle, and an outer sealing ring is arranged between the hand wheel base and the outer adjusting handle. The mechanism has the advantages of being easy to assemble, good in stability, convenient to use and high in precision.

The invention belongs to the calculation holographic three-dimensional display field and particularly relates to a colorful holographic three-dimensional display method based on space division and a system thereof. The system utilizes a CCD camera or a computer graphical method to acquire a multi-perspective two-dimensional image of a colorful target object, through pre-processing transformation, a set of target images suitable for phase type hologram calculation are generated; an iteration Fourier transform algorithm is utilized to calculate corresponding phase type holograms; a single spatial light modulator is divided into red, green and blue sub zones, the sub zones respectively load corresponding red, green and blue component holograms, a hologram diffraction reproduction characteristic is utilized to generate three single-color holographic reproductive images on a reconstruction plane, directional diffraction of pixels of the three reproductive images is carried out to a fixed position to form different view points, multi-perspective colorful reproductive images are acquired, and true colorful three-dimensional display is realized; in combination with a spatial light modulator, rapid refresh loading of the holograms is carried out, and multi-perspective true colorful full-parallax dynamic holographic three-dimensional reproduction is realized.