9results about How to "Reduce contrast" patented technology

The invention discloses a road crack segmentation method and device based on deep learning, electronic equipment and a program product. According to the method, two collaborative and functional complementary processing paths are adopted to perform parallel processing on a to-be-segmented image so as to obtain corresponding feature information: on one hand, the to-be-segmented image is converted into a feature sequence containing spatial position information, and global semantic modeling is performed on the feature sequence; global semantic features representing the overall shape of the crack and long-range semantic dependence are obtained; and on the other hand, local texture and edge detail information of the crack is directly extracted from the to-be-segmented image, and detail enhancement features are obtained. Thirdly, fusing the global semantic features and the detail enhancement features to obtain target fusion features considering crack positioning, connectivity and boundary and texture expression; according to the segmentation result output based on the target fusion features, even tiny cracks can be effectively depicted, and then the continuity, boundary definition and robustness of tiny crack segmentation under the complex road background can be improved.

The utility model relates to the technical field of optical lenses, in particular to a miniature dot matrix contrast-reducing lens, which comprises a lens, the lens comprises a central optical area and a miniature dot matrix area, the miniature dot matrix area comprises two dot matrix structures, namely a large dot matrix with the diameter of 0.18 mm and a small dot matrix with the diameter of 20-30 microns, the number ratio of the large dot matrix to the small dot matrix is 1: 3-1: 8, and the number ratio of the large dot matrix to the small dot matrix is 1: 3-1: 8. The covering density of the dot matrix areas is 10-45%, the dot matrix spacing is 0.2-0.65 mm, the depth of the dot matrix is 2-8 microns, the edge of the dot matrix is provided with a smooth transition area of 0.05-0.1 mm, the lens further comprises micro-lens areas, and the micro-dot matrix areas and the micro-lens areas are alternately distributed. The central optical area of the lens is not covered by the dot matrix, so that clear view is guaranteed; the contrast of the dot matrix area is accurately regulated and controlled, full-view blurring is avoided, micro-lens defocusing and dot matrix contrast reduction are combined, eye axis growth is doubly inhibited, dot matrix size and interval optimization is achieved, and glare and visual interference are reduced.

The present disclosure provides a light-emitting device with improved light output and a preparation method thereof, and belongs to the technical field of optoelectronic manufacturing. The light-emitting device comprises a light-emitting functional layer, a driving substrate and a brightening functional layer. The light-emitting functional layer is located on the driving substrate, and comprises a plurality of light-emitting units arranged at intervals. The brightening functional layer comprises a bearing layer, a reflective layer and a plurality of light-emitting lenses. The bearing layer is located on the surface of the light-emitting functional layer away from the driving substrate. The surface of the bearing layer has a plurality of light-emitting grooves exposing the light-emitting functional layer. Each light-emitting groove exposes at least one light-emitting unit. The reflective layer is located on at least the groove wall of each light-emitting groove. The light-emitting grooves correspond to the light-emitting lenses one by one, and the light-emitting lenses are located in the corresponding light-emitting grooves. The embodiments of the present disclosure can improve the problem of low light output of the light-emitting device and improve the brightness of the light-emitting device.

The application discloses a medical image tumor segmentation model based on a TransUNet framework and a construction method and a segmentation method thereof, the construction method comprising constructing a segmentation model based on the TransUNet, wherein an encoder of the segmentation model comprises CNN modules and a self-defined VSS module connected in sequence, the output ends of each convolution block of the CNN modules are input into a decoder of the segmentation model through a jump connection layer, the self-defined VSS module comprises a plurality of novel conversion layers stacked in series based on a Mamba module and a focus linear attention module, the self-defined VSS module performs multi-level feature interaction and global modeling on first encoding features output by the CNN module to obtain second encoding features; and the decoder performs feature fusion and up-sampling on the second encoding features and the output of the jump connection layer to obtain a segmentation image for a sample image. The application can significantly improve the segmentation robustness and precision of breast ultrasound images with low contrast, high noise and fuzzy boundaries.

The invention belongs to the technical field of visual inspection, and particularly relates to electroplated product online image detection equipment which comprises detection equipment, a conveying table arranged in the detection equipment and used for conveying electroplated products and a detection head arranged in the detection equipment. An air inlet is formed in the end, away from the detection head, of the protective cover, the protective cover is externally connected with a negative pressure fan through an air pipe, a treatment box is arranged in the protective cover and communicates with an inner cavity of the protective cover through an air hole, a filter plate is fixedly installed in the treatment box, and a storage box is rotationally installed at the bottom end of the filter plate; a wiping roller attached to the bottom of the filter plate is movably mounted in the storage box and is hollow, a liquid outlet is formed in the outer wall of the wiping roller, corrosive mist and metal dust in air are sucked and prevented from being attached to the surfaces of a lens and a light source of the detection head, imaging blurring and contrast reduction are prevented, and continuous and stable operation of the detection process is guaranteed.

This utility model discloses a screen structure and a high-gain white screen, belonging to the field of projection display technology. The screen structure includes two substrate layers, with a barrier layer bonded between the two substrate layers via an adhesive layer. The substrate layers, adhesive layer, and barrier layer have different refractive indices. The high-gain white screen includes a protective layer, a diffusion layer, and the aforementioned screen structure, arranged sequentially from the incident surface to the back surface. By utilizing the difference in refractive indices between the substrate layer, adhesive layer, and barrier layer, an optical effect similar to mother-of-pearl is created, reducing light penetration and enhancing light reflection and scattering. This improves detail rendering, reduces the loss of dark areas, enhances resistance to external light interference, reduces contrast, reduces color saturation distortion, reduces light loss during propagation, improves overall brightness uniformity, ensures full presentation of high-resolution signal details, and is compatible with more projection devices.

The invention discloses a light source device and an optical imaging system. The first laser light source comprises a first laser module and a second laser module, the first laser module is used for generating first laser, and the second laser module is used for generating second laser. The second laser light source comprises a third laser module and a fourth laser module, the third laser module is used for generating third laser, and the fourth laser module is used for generating fourth laser. The first polarization light combination module is arranged on a light path where the first laser and the third laser are located and used for generating first mixed light. The second polarization light combination module is arranged on the light path where the second laser and the fourth laser are located and used for generating second mixed light. The wavelength light combination module is arranged on a light path where the first mixed light and the second mixed light are located. According to the light source device in the embodiment, light combination is conducted on multiple paths of laser in the mode that polarization light combination is conducted firstly and then wavelength light combination is conducted, light spots of finally generated emergent light can be small, and the design difficulty and hardware cost of elements in a follow-up light path are reduced.

This utility model discloses a material transport system with integrated status monitoring and energy-saving functions, belonging to the technical field of material transport equipment. It includes an upper-level terminal, a lower-level controller, and at least three belt conveyors connected in sequence. All belt conveyors are connected to the lower-level controller, which in turn is connected to the upper-level terminal. Each belt conveyor's drive motor is equipped with an integrated temperature and vibration sensor to detect abnormal vibration and temperature. High-definition cameras are installed under the belts to detect belt tears. A material flow detection device is installed above the belts to detect material passing through a designated location and control the start and stop of the next-level belt conveyor accordingly. This utility model effectively solves the problems of high energy consumption during belt conveyor startup, low transport efficiency, and insufficient equipment health status monitoring in existing material transport systems.

The application relates to a video image processing method and device, electronic equipment and a storage medium. The method comprises the following steps: performing background and target separation processing on multiple images in a target video respectively to obtain multiple target background images and multiple first target images, counting the total number of pixel points in the multiple first target images and the number of times that each first pixel point in the multiple first target images appears in the multiple first target images; based on the ratio of the total number of pixel points to the number of times that each first pixel point appears, second pixel points that do not satisfy a probability threshold range are selected from each first pixel point in the multiple first target images; the pixel information of the second pixel points in the first target image is replaced by the pixel information corresponding to the pixel coordinates of the second pixel points in a preset reference image to obtain a second target image; and the second target image and the corresponding target background image are merged to obtain multiple target images. The video quality can be effectively improved.