94 results about "Corner detection" patented technology

The invention relates to the technical field of image processing, in particular to an online calibration method and system for hydrological flow data, and solves the technical problem of low flow calculation precision caused by rough window segmentation and illumination interference in the prior art. The method comprises the following steps: collecting a water surface video in a river water flow process, and setting a velocity measurement line area in the water surface video along a water flow direction; according to the gray level distribution of the velocity measurement line area, carrying out illumination state judgment, and according to a judgment result, carrying out image restoration on an illumination abnormal area in the water surface video to obtain a restored water surface video; constructing a space-time image for the repaired water surface video, performing corner detection, clustering corners to divide a plurality of windows, and constructing a window weight for each window; and weighting the texture angle of each window according to the window weight of each window to obtain a weighted texture angle, and calculating hydrological flow data according to the weighted texture angle.

The invention discloses an acupoint recognition and positioning method and system, a moxibustion robot and a storage medium. The method comprises the steps of obtaining multi-modal data of a human body to be subjected to physiotherapy; carrying out preprocessing on the multi-modal data; constructing a multi-modal feature fusion model, and generating an enhanced acupoint feature map; constructing a human body meridian-acupoint incidence relation matrix on the basis of the enhanced acupoint feature map in combination with a preset human body meridian-acupoint incidence relation matrix; dynamically compensating the candidate area, calculating the real-time offset of the acupuncture point based on the body surface micro-motion data, and correcting the depth coordinate of the acupuncture point in combination with the body surface elastic modulus data to obtain the accurate three-dimensional coordinate of the acupuncture point to be positioned; on the basis of the precise three-dimensional coordinates, blood vessels and nerve dense areas around the acupuncture points are recognized through an improved Harris corner detection algorithm, and a safe moxibustion applying path is generated; meanwhile, the confidence value of the precise three-dimensional coordinate is calculated. According to the method, high-precision, self-adaptive, all-position compatible, safe and controllable acupuncture point recognition positioning is realized.

The invention relates to the technical field of industrial detection, in particular to a pin defect detection method, device and equipment and a medium. According to the method, based on pin target positioning, image segmentation is carried out on a pin image, form extraction is carried out on a pin area, an actual included angle between a central axis and a reference axis is calculated, whether the actual included angle is in a preset angle range is detected, and if the actual included angle is detected not in the preset angle range, the pin area is determined. And if so, determining that the corresponding pin has a bending defect. The method comprises the following steps: acquiring a pin image from an original image based on pin target positioning, segmenting to obtain a pin area, extracting morphological characteristics of the pin area to determine a central axis, determining a reference axis according to the pin image, and calculating an actual included angle between the reference axis and the pin image; and finally, determining whether the pin has a bending defect or not by judging whether the included angle is within a preset range or not. And performing morphological analysis on the pin image to determine an included angle between the central axis and the reference axis so as to complete defect detection of the pin.

The utility model belongs to the technical field of detection rulers, and particularly discloses a pressure vessel corner angle detection ruler. Comprising a main ruler and an auxiliary ruler which are rotationally connected, indicating mechanisms and level bubbles are arranged on the main ruler and the auxiliary ruler, the indicating mechanisms are located at the free ends of the main ruler and the auxiliary ruler, and a detection mechanism is movably arranged on the front side of the main ruler. The main ruler and the auxiliary ruler are rotationally connected, so that the total length of the detection ruler is flexibly adjusted, the auxiliary ruler can rotate around the rotating shaft relative to the main ruler, integrated detection of multiple parameters such as length, straightness, curvature, height and width can be realized, and the detection efficiency is greatly improved. The problem that the existing corner angle detection technology is difficult to adapt to the actual requirements of portability, instantaneity, manual operation and direct reading of specific deviation values in the manufacturing, mounting and inspection field of the pressure vessel is solved.

The invention discloses a method and a system for quickly inputting voting information. The method comprises the following steps: shooting a paper voting table containing a two-dimensional code to obtain an original image; performing document four-corner detection on the original image, and performing perspective transformation processing to obtain a coarse correction image; analyzing the two-dimensional code and pulling the corresponding electronic template; key points are extracted from the rough correction image and the template image for matching, and an initial key point pair set is obtained; when the quality is insufficient, the key point pairs are supplemented through OCR character graph matching; establishing an adaptive FFD control grid, and carrying out nonlinear deformation on the rough correction image until the rough correction image is aligned with a template pixel level to obtain a fine correction image; and clipping the ROI region according to the template definition file, executing check box classification identification to obtain structured voting data, and packaging the structured voting data into a preset format and outputting the structured voting data. According to the method, the heterogeneous second-level, high-robustness and traceable automatic information input aiming at the paper voting table form is realized.

The invention discloses a measuring instrument inclination angle monitoring method and device fused with vision assistance, belongs to the technical field of precise instrument measurement, and aims to solve the problem of poor reliability of inclination angle detection of a measuring instrument in the prior art. According to the technical key points, the method comprises the following steps: respectively acquiring an image containing a level gauge and an inclination angle measurement value by using an image sensor module and an inclination angle sensor in an inclination angle detection device fixed on a measuring instrument; a multi-core processor in the dip angle detection device is used for processing images containing the level gauge, bubble position information of the level gauge is extracted, and a dip angle value of the level gauge is obtained according to the bubble position information; and carrying out fusion processing on the dip angle value of the level gauge and the dip angle measurement value to obtain a fused dip angle value, and taking the fused dip angle value as a final dip angle monitoring result of the measuring instrument. The method is suitable for a high-precision measuring instrument with extremely high requirements on the real-time performance and the accuracy of horizontal inclination angle monitoring.

The application relates to the technical field of image recognition, and particularly provides an unmanned aerial vehicle part automatic recognition method based on corner point detection, which comprises the following steps: constructing a neural network model comprising a channel attention mechanism, acquiring multiple unmanned aerial vehicle images, and pre-processing and unmanned aerial vehicle part labeling the unmanned aerial vehicle images; acquiring corner point information in the unmanned aerial vehicle images after the unmanned aerial vehicle parts are labeled through a corner point detection method, and generating a corner point feature map; fusing the corner point feature map and the unmanned aerial vehicle images after the unmanned aerial vehicle parts are labeled to obtain a corner point image; and training the neural network model again by using the corner point image, inputting a to-be-detected target image into the trained neural network model, and recognizing unmanned aerial vehicle parts in the to-be-detected target image. The application enhances the robustness of the neural network model to complex scenes and noise interference, and realizes accurate and efficient unmanned aerial vehicle part automatic recognition.

The invention discloses a robot for detecting in a bridge box cavity based on a flexible cable track and an operation method, and belongs to the technical field of bridge detection equipment. Comprising a mobile platform, a walking assembly, an anti-collision safety assembly, a detection assembly, a distance control assembly and a control system, a steel cable serves as a flexible track, the walking assembly clamps the steel cable to move through a V-shaped walking wheel and a safety lock rod, double encoders are fused for positioning, and a multi-degree-of-freedom mechanical arm of the detection assembly drives a detection unit to collect data in a full-dimensional mode. And anti-collision and self-adaptive lighting functions are matched. According to the invention, no-dead-corner detection of the box cavity is realized, the flexible track is convenient and reusable in deployment, accurate in positioning, safe and efficient, manual risks are avoided, the detection efficiency and data reliability are greatly improved, and the method is suitable for detection of various bridge box cavities.

The application provides a certificate integrity detection method and device and a storage medium. The method comprises the following steps: performing edge extraction on a certificate image based on a double-flow convolutional neural network (GSCNN) model to obtain an edge image of the certificate, and / or performing edge corner classification on an edge corner image of the certificate based on an edge corner detection network model to obtain an edge corner detection result; the edge image is used to determine whether the edges of the certificate belong to a closed loop; the edge corner detection result is used to determine whether the edge corners of the certificate are complete; and in the case that the edges of the certificate belong to a closed loop and the edge corners of the certificate are complete, it is determined that the certificate is complete. The certificate integrity detection method and device and the storage medium provided by the application determine whether the certificate is complete according to two automatic detection results of edge detection and edge corner detection, improve the accuracy of certificate integrity determination, improve the certificate acquisition quality, and reduce the manpower and material costs of certificate detection.

The invention discloses microfocus X-ray visual inspection equipment for surface defects of a plate-shaped object. An objective table of the microfocus X-ray visual inspection equipment is used for horizontally bearing the detected plate-shaped object; the radiation source is a signal emitter, and the detector is a signal collector; the objective table is used for achieving translation in the Z direction and the X direction and can achieve rotation in the horizontal plane, and the radiation source and the detector which are located above the objective table and below the objective table respectively can swing along with the rectangular frame and can move on the Y-direction rod in the Y direction. Three-dimensional dead-corner-free detection is achieved through multi-axis linkage, the definition and detection precision of acquisition imaging are effectively improved, the phenomena of missing detection and false detection are prevented, and the device is particularly suitable for 2D and 3D detection of chip packaging defect micro-focus X-ray vision.

The present application relates to the technical field of data processing, in particular to a data processing method and system based on unmanned aerial vehicle video and vector data fusion. By calculating the deviation of the predicted position and the actual detection position of each reference point in the image, the reference points with large errors are screened and the camera pose parameters are adjusted using the random consistency algorithm and the iterative optimization method to improve the projection accuracy and stability. Through grid division, corner detection and feature descriptor extraction, a reference point set is established, the parallax change standard deviation is introduced, and the unstable reference points with large fluctuations are removed to ensure the reliability of the pose solving process and the consistency of spatial registration. According to the registration and fusion of unmanned aerial vehicle video and vector geographic data, the problems of current vector element drift, occlusion determination difficulty and attitude error dynamic correction are solved, the spatial reality and dynamic stability of the augmented reality scene are enhanced, and the industrial needs of various application scenarios such as underground pipe network inspection are met.

The application discloses a kind of based on machine vision's preceding group subassembly weld joint positioning system and method, comprising the following steps: camera is hoisted above welding platform by tooling, relative position is fixed and camera center line is perpendicular to the plane of the workpiece to be measured;Camera intrinsic calibration is carried out, and hand-eye calibration between camera-welding robot or hand-eye calibration between line laser sensor-welding robot;Workpiece feeding, scene picture is shot;Based on the shooting picture, workpiece feature point set calculation is carried out using gray scale segmentation and Harris corner detection, for the scene that needs to be shot by multiple cameras, image stitching based on calibration board guide map is carried out before workpiece feature point set calculation;Seeking location feature point set and calculating weld joint end point;Weld joint tracking is carried out, and weld joint data is corrected in real time, to obtain actual weld joint data.The application can effectively improve the welding efficiency and automation degree of shipyard small group subassembly unit.

The invention provides a mineral dynamic identification and accurate sampling method and system based on deep learning, and the method comprises the steps: obtaining mineral image data and laser radar ranging data on a mine car in an underground environment, and carrying out the preprocessing of the data, and obtaining mineral feature data; designing a deep convolutional neural network architecture with multi-scale feature fusion and performing multi-stage training to determine a mineral recognition model; performing target detection by using a mineral recognition model, and realizing multi-target tracking through a Kalman filter and a Hungary algorithm; calculating mine car motion parameters by using FAST corner detection, ORB descriptors and an optical flow algorithm; predicting the future position of the target mineral by using multi-sensor fusion and an extended Kalman filter; and based on the predicted position, precise sampling operation is executed through quintic polynomial interpolation and dynamic compensation control. According to the invention, high-precision mineral identification and dynamic accurate sampling can be realized in a complex underground environment, and the representativeness and production efficiency of mineral sampling are obviously improved.

The utility model relates to the technical field of medicine automatic detection and information checking, in particular to a medicine box traceability code reading and information checking device integrating all-dimensional visual angle detection. The device comprises a conveying unit, a fixed support fixedly arranged on the conveying unit, a top detection camera, a bottom detection camera, a left detection camera, a right detection camera, a front detection camera, a rear detection camera and a central processing unit. The beneficial effects of the utility model are that according to the technical scheme, the detection cameras are arranged in the upper, lower, left, right, front and back directions of the conveying unit, so that omnibearing visual angle detection of six surfaces of the medicine box is realized, and the problem of traceability code information omission caused by fixed installation positions of the cameras in the prior art is solved.

The application discloses a wafer chamfer detection method and computer equipment, and relates to the technical field of LED, which comprises the following steps: acquiring the reference position of each wafer in a wafer box; driving a handling hand to send the wafer to a first vacuum adsorption table of a center positioning mechanism to position the wafer; driving the handling hand to send the wafer to a second vacuum adsorption table of an edge face width measuring mechanism; driving the second vacuum adsorption table to rotate to acquire image information of the sidewall of the wafer in the rotating process, and calculating the face width of the flat edge position and the circumferential position of the wafer according to the image information; driving the handling hand to send the wafer to a third vacuum adsorption table of a diameter measuring mechanism; driving the third vacuum adsorption table to rotate to measure the distance between the flat edge position and the center of the wafer, the maximum diameter and the minimum diameter; and driving the handling hand to move the wafer to the corresponding reference position of the wafer. The application can effectively improve the measurement speed and the measurement accuracy.

A grating image-based house layout identification and three-dimensional reconstruction method and system belong to the technical field of house layout image identification.The present application comprises: step 1, collecting and screening high-quality layout maps suitable for the national conditions of China to construct a grating layout map vector dataset; step 2, identifying the walls, doors and windows and scale in the layout map based on a key point detection network, generating and screening candidate primitives through axial alignment rules and constraints; step 3, detecting and positioning the scale end points in a Yolov8 and Shi-Tomasi corner detection combined manner, and recognizing the scale digital characters through a pre-trained multi-modal optical character recognition model OFA-OCR; step 4, designing and implementing a layout identification and three-dimensional reconstruction system based on a WebGL framework, supporting the rendering of the walls and doors and windows after reconstruction, the user interaction with the reconstructed layout, the forward and backward functions of the layout operation history record.

The invention provides an encodable liquid crystal image processor, a preparation method and a broadband parallel imaging device. The encodable liquid crystal image processor comprises a first substrate, a second substrate and a cholesteric liquid crystal layer which are oppositely arranged, one side of the first substrate is provided with a first orientation layer, and one side of the second substrate is provided with a second orientation layer; the first orientation layer and the second orientation layer comprise control patterns, and the control patterns enable liquid crystal directors to meet orientation angle distribution so as to guide liquid crystal distribution in the cholesteric liquid crystal layer; in a wave band covered by a chiral photonic band gap of the cholesteric liquid crystal layer, space parallel image processing with multiple functions is realized at the same time. According to the invention, a multi-task image processing function, such as edge detection and corner detection, is integrated on a single element, and spatial parallel output is realized in a visible light wave band by utilizing the broadband geometric phase modulation characteristic of cholesteric liquid crystal, so that the limitations of single function and narrow bandwidth of an existing image processing device are effectively improved.

The invention discloses a scanner equipment image paper folding angle detection method and system, a medium and a product, and relates to the technical field of image processing. According to the method, firstly, a scanning image is obtained and converted into a grey-scale map for edge detection, and a boundary point set is extracted through bidirectional scanning; and performing region division on the boundary point set, extracting corner region boundary points, constructing an angle histogram, screening a dominant angle sequence, and outputting a first break angle detection signal. And if no break angle is detected, straight line detection is carried out to determine grouped straight lines, a base line is selected to carry out rotation correction, and the paper boundary is determined and cutting is carried out. And extracting a corner region sub-image from the cut image, and performing convex polygon fitting to output a second break angle detection signal. And finally, determining a final break angle detection result according to the two detection signals. By implementing the technical scheme provided by the invention, the accuracy of paper folding angle detection can be improved.

The utility model relates to an appearance detection device for a PAD back shell. The appearance detection device comprises a feeding and cleaning machine, an AOI detection machine and a discharging machine, the feeding machine comprises a feeding flow channel assembly, a large face cleaning assembly, a carrying assembly, a side face and R angle cleaning assembly and an overturning assembly. The AOI detection machine comprises a feeding carrying assembly, a three-rotor carrying assembly, a system plane detection assembly, a side face and R angle detection assembly and a user plane detection assembly. The three-rotor carrying assembly comprises a first rotor carrying assembly, a second rotor carrying assembly and a third rotor carrying assembly. The cleaning machine is arranged in the device, so that large surfaces, side surfaces and R angles of products can be cleaned, and the detection rate of the products is ensured; the detection equipment comprises system surface detection and side surface amp; r angle detection and user plane detection are carried out, and each visual detection assembly comprises a plurality of groups of cameras and is used for carrying out omnibearing detection on six surfaces of a product; product smudginess can be reduced, and interference of smudginess on visual inspection is reduced; and the standardization of the machine is realized, the production cost is reduced, and the production efficiency is improved.

The invention relates to the technical field of equipment belt detection, discloses a conveying belt longitudinal tear detection method and equipment based on linear laser visual analysis, and is used for solving the problems of detection lag and high false detection rate of an existing tear detection scheme. The method comprises the following steps: acquiring a belt surface image when linear laser equipment is projected to a conveying belt; extracting laser line points in the belt surface image by using a laser line detection model taking ResNet18 as a backbone network, and fitting all the extracted laser line points to obtain a laser curve image; identifying a first tearing candidate area in the laser curve image by using a preset corner detection algorithm, and carrying out differential analysis on the belt surface image and the laser curve image to obtain a second tearing candidate area; according to the first tearing candidate area and the second tearing candidate area, z-score is used for carrying out tearing analysis, and a belt tearing detection result is obtained.

The utility model discloses a verticality detection device which comprises a right-angle detection frame, one side of the right-angle detection frame is connected with two limiting pipes in a penetrating mode, the limiting pipes are connected with a detection rod in a penetrating mode, one end of the detection rod is fixedly connected with a limiting reflecting plate, and the other end of the detection rod is fixedly connected with a sliding seat. The beneficial effects of the utility model are that the moving right-angle detection frame slowly approaches to the wall surface to be detected, so that the movable detection rod on the right-angle detection frame is attached to the wall surface, at the moment, the pressed detection rod can move on the limiting pipe, and the detection rod does not move in the moving process. The infrared laser range finders arranged on the other side of the right-angle detection frame can detect the distance between the limiting pipes, when the distances detected by the two infrared laser range finders are kept consistent, the wall surface is vertical, otherwise, the wall surface is inclined, the structure is reasonable, the device is convenient to move and detect, and the perpendicularity of the wall surface can be rapidly detected.

A hitch angle detection device detects a trailer hitch angle based on a trailer image shot by a vehicle camera using a model obtained by performing learning using learning data of a learning trailer image shot by a learning vehicle camera and a label indicating learning trailer hitch angle, additional learning of the model is performed in a server by using additional learning data of an additional learning trailer image shot by the vehicle camera and the label indicating the trailer hitch angle when the additional learning trailer image is shot, input of the trailer hitch angle is accepted by a vehicle HMI, the hitch angle detection device detects the trailer hitch angle based on the trailer image shot by the vehicle camera by using the model with the additional learning performed in the server.

The application discloses a kind of target ground object AI identification methods based on multi-source aerial photograph, it is related to aerial photograph processing technical field, including the following steps: S1, the panorama aerial photograph of the region to be identified is collected;S2, the first data set is extracted using the gray level co-occurrence matrix of panorama aerial photograph;S3, according to the response value of all corner points in panorama aerial photograph, the second data set is extracted;S4, the first data set and the second data set are all input to the preliminary screening condition model, and the pixel point meeting the preliminary screening condition model is regarded as the suspected target ground object of the region to be identified;S5, the suspected target ground object of the region to be identified is processed, and the accurate target ground object is obtained.The application is to overcome the limitation of single feature, and multiple information sources are fused, such as the texture features extracted by gray level co-occurrence matrix and the structural features extracted by corner detection, which can more comprehensively describe the characteristics of image control points in aerial photography, improve the accuracy and robustness of identification.

The invention relates to the technical field of chip detection, in particular to a chip defect detector based on multispectral imaging, which comprises a detector main body, a discharge port is arranged at the center of the front surface of the detector main body, an adjustable fixing mechanism is arranged in the detector main body, and an image taking mechanism is arranged above the fixing mechanism. According to the invention, by optimizing a fixing and image capturing structure and combining a multispectral imaging and intelligent analysis system, all-directional dead-corner-free detection of the chip is realized. The multi-dimensional adjustment capability of the fixing mechanism guarantees the detection stability and comprehensiveness, and the spectrum switching and light path design of the image capturing mechanism improves the defect identification precision. The equipment integrates an automatic process and an intelligent analysis function, greatly reduces the omission ratio and manual intervention, is compatible with chips of various specifications, gives consideration to the detection efficiency and the refinement level, provides an efficient and accurate solution for chip manufacturing quality management and control, and assists in improving the reliability and market competitiveness of chip products.

Embodiments include a system for mobile check deposit enabling auto-capture functionality via video frame processing. The system includes a mobile device with a camera, processor, and memory storing instructions that, when executed by the processor, configure the mobile device to generate a live video of a check in the camera's field of view, monitor the live video with respect to at least one monitoring criterion, and automatically capture a frame of the live video when that frame satisfies the monitoring criterion. The monitoring criterion may be based on light contrast, light brightness, check positioning, check dimensions, character spacing, skewing, warping, corner detection, edge detection, and MICR line detection. The mobile device may be configured to confirm that mobile check deposit can go forward after optical character recognition (OCR) is performed on the check, a MICR line of the check is recognized, and an amount of the check is validated.

This invention discloses an event-visual-inertial semantic simultaneous localization and map building method, which is applicable to navigation and localization of mobile robots, unmanned vehicles and drones. This method acquires event data streams from event cameras and inertial data from IMUs, and optionally standard camera images. Based on the event data, it constructs an event activity surface, performs coarse-to-fine event corner detection on it, constructs an event representation for event optical flow estimation, and estimates the event optical flow. It then uses the event optical flow to perform temporal tracking of event corners to obtain multi-time-stack related observations. Based on the event corners and event representations, it extracts descriptive information, performs loop closure detection and relocalization, and introduces loop closure constraints as additional residual terms into a sliding window graph optimization. Within the sliding window, it combines residuals from events, images, IMUs, edge detection, and loop closure relocalization, and sets adaptive weights for multi-source fusion to obtain a six-degree-of-freedom pose sequence. It performs target detection on the event data stream, outputting detection boxes, categories, and confidence scores. Based on the event optical flow, it implements missed detection compensation and motion consistency checks to identify dynamic target regions, eliminates dynamic corners and their observation constraints, and generates an object-level semantic map based on pose estimation. This improves the robustness of localization mapping and environmental understanding in complex lighting, high-speed motion, and dynamic interference scenarios.

The application belongs to the technical field of computer vision, and discloses a QR calibration board calibration method and system for multi-scene cooperation, which comprises the following steps: constructing a fusion calibration board, adopting a unified two-dimensional matrix structure, integrating multiple functional areas in the same calibration board, and cooperating with layout according to preset space rules; generating a positioning mark area and an adjacent formatted information area in the two-dimensional matrix structure; configuring the physical size of the positioning mark area with a fixed internal structure, and storing format constraint information through the formatted information area; configuring a visual corner detection pattern in the calibration board according to different application scenarios, and performing structural completion on the corner area affected by the two-dimensional matrix layout; writing preset task information into the data content area in the multi-functional area, and generating a code word based on a finite field coding rule before writing; and the QR calibration board has improved adaptability and calibration precision in multi-scene cooperative application.

The application discloses a kind of double coordinate system bolt positioning and angle detection method based on space and fan modeling, comprising: step 1, establish bolt database;Step 2, establish the double coordinate system model of associated space coordinate system and fan modeling coordinate system, establish the coordinate conversion between double coordinate system, digitize the bolt to be detected, bind its accurate coordinate;Step 3, based on target detection network, construct bolt target detection model;Step 4, the image collected by unmanned aerial vehicle is perspective correction processing, based on edge detection identification bolt center point, based on deep learning algorithm realizes bolt feature detection and feature matching, calculates bolt rotation angle, realizes bolt angle detection.The application establishes bolt database and double coordinate system model systematically before detection operation starts, carries out centralized digital pretreatment and integration to bolt position, state and environmental factors, forms high-precision operation guide and reference information that can be flexibly called by unmanned aerial vehicle detection system.

This application relates to the field of image detection technology, and provides a corner detection method, system, electronic device, and medium based on texture features. The method includes: extracting texture features from an image to be processed to obtain a texture image; fusing the texture image with the image to be processed to obtain a fused image; and performing corner detection on the fused image based on the texture image. This method separates the background and texture of the image to be processed to obtain a texture image, fuses the texture image with the image to be processed to obtain a fused image, and performs corner detection on the fused image based on the texture image. By using this method to detect corners based on the texture features of the image, the selected pixel positions can better reflect the corner detection performance, greatly saving computational power and time.

An image processing method includes the steps of: performing corner detection on a specific pixel in image data using a plurality of corner detection filters to generate a plurality of detection results, wherein the plurality of corner detection filters respectively correspond to a plurality of corners having different directions; determining which one of the plurality of corners having different directions the specific pixel belongs to according to at least the plurality of detection results; and performing image processing corresponding to a specific corner of the plurality of corners to which the specific pixel belongs on the specific pixel according to the specific corner to generate processed image data.