Abstract: The invention provides an image segmentation method and an electronic device. The image segmentation method includes the following steps. Regression analysis is performed on a first gray-scale image to obtain a residual image having an object backbone area. A pixel value of each pixel in the object backbone area is defined as an average gray-scale value of the object backbone area in the residual image, and a second gray-scale image having the object backbone area is generated. It is recursively determined whether a residual polarity of each adjacent pixel adjacent to edge pixels of the object backbone area in the residual image is the same as a residual polarity of the corresponding edge pixel, and whether a pixel value of each adjacent pixel is greater than a first threshold, so as to expand the object backbone area in the second gray-scale image, which is extracted as a target object.

Abstract: An automated moving vehicle and a control method thereof. The control method includes: obtaining a static point cloud and a current point cloud on a work space through a sensor, wherein the current point cloud includes a current scan point; calculating a shortest distance between the current scan point and the static point cloud; in response to the shortest distance being greater than or equal to a first threshold, calculating a first distance between the current scan point and the automated moving vehicle; and in response to the first distance being less than a second threshold, controlling a driving device of the automated moving vehicle to stop a movement of the automated moving vehicle.

Abstract: A supervised training device and method for training a neural network, and a supervised classification method and device based on the neural network are provided. The training device includes a storage unit and a processor. The processor accesses and executes a sampling module, a labelling module and a training module in the storage unit. The sampling module samples a first image and a second image from a first dataset. The labelling module tags the first image in response to a first control instruction to generate a first tagged image, and generates label data according to the first tagged image and the second image. The training module trains the neural network according to the label data.

Abstract: A method and an electronic apparatus for stitching a three-dimensional spherical panorama are provided. The method includes: obtaining a first image and a second image; projecting the first and seconds image onto a virtual spherical plane to form a first equirectangular panorama and a second equirectangular panorama; duplicating the first equirectangular panorama to generate a third equirectangular panorama, and duplicating the second equirectangular panorama to generate a fourth equirectangular panorama; performing image enhancement on the third and fourth equirectangular panoramas to generate a first enhanced equirectangular panorama and a second enhanced equirectangular panorama; calculating an optical flow for the first and second enhanced equirectangular panoramas; generating the three-dimensional spherical panorama according to the first and second equirectangular panoramas and the optical flow.

Abstract: An electronic device having a neural network framework for estimating optical flow is provided. The electronic device is connected to an image acquiring unit, which acquires images to be analyzed. The electronic device includes a storage unit, a feature extraction unit, an optical flow estimation unit and a refining unit. The storage unit stores a feature extraction module. The feature extraction unit is connected to the image acquiring unit and the storage unit. The optical flow estimation unit is connected to the feature extraction unit to generate an estimated optical flow. The refining unit is connected to the optical flow estimation unit to input the estimated optical flow to a refining module to obtain an estimated optical flow result. A method for estimating optical flow is also provided to reduce the number of training parameters required for estimating optical flow, thereby reducing a training time and improving training stability.

Abstract: An object positioning system and an object positioning method are provided. The object positioning system includes a sensing device, a storage device and a processing device. The sensing device collects point cloud data obtained from a scene including a target object. The processing device inputs surrounding area data centered on a key point and a preset feature descriptor to a neural network to calculate a scene feature descriptor of the scene. The processing device performs feature matching between the scene feature descriptor and the preset feature descriptor, and calculates a position of the target object in an actual space. The invention utilizes the feature extraction capability of the neural network to effectively improve the accuracy and stability of target object identification and positioning.

Abstract: A method for filtering periodic noise and a filter using the method are provided. The method includes: obtaining an input signal; detecting a fundamental frequency corresponding to a maximum peak in a spectrum of the input signal, detecting a harmonic frequency according to the fundamental frequency, and detecting an aliasing frequency corresponding to the harmonic frequency in response to the harmonic frequency corresponding to the fundamental frequency being greater than a Nyquist frequency of the input signal; filtering the fundamental frequency and at least one of the harmonic frequency and the aliasing frequency of the spectrum to generate a first filtered spectrum, and restoring the input signal according to the first filtered spectrum to generate an output signal; and outputting the output signal. The method for filtering the periodic noise and the filter using the method may filter the periodic noise in the input signal affected by an aliasing effect.

Abstract: A method and a system for inspecting a display image are provided. A test image is displayed on a display surface through a display device. An optical inspection image is generated by photographing the test image on the display surface by an image capturing device. A brightness channel image and a color channel image of the optical inspection image are obtained by performing color space transformation on the optical inspection image. Background estimation is applied to the color channel image to obtain a color channel background image. At least one color nonuniformity region in the test image is obtained by respectively comparing an inspection reference value with multiple color component pixel values of the color channel background image.

Abstract: A method and an electronic apparatus for stitching a three-dimensional spherical panorama are provided. The method includes: obtaining a first image and a second image; projecting the first and seconds image onto a virtual spherical plane to form a first equirectangular panorama and a second equirectangular panorama; duplicating the first equirectangular panorama to generate a third equirectangular panorama, and duplicating the second equirectangular panorama to generate a fourth equirectangular panorama; performing image enhancement on the third and fourth equirectangular panoramas to generate a first enhanced equirectangular panorama and a second enhanced equirectangular panorama; calculating an optical flow for the first and second enhanced equirectangular panoramas; generating the three-dimensional spherical panorama according to the first and second equirectangular panoramas and the optical flow.

Abstract: The invention provides an image segmentation method and an electronic device. The image segmentation method includes the following steps. Regression analysis is performed on a first gray-scale image to obtain a residual image having an object backbone area. A pixel value of each pixel in the object backbone area is defined as an average gray-scale value of the object backbone area in the residual image, and a second gray-scale image having the object backbone area is generated. It is recursively determined whether a residual polarity of each adjacent pixel adjacent to edge pixels of the object backbone area in the residual image is the same as a residual polarity of the corresponding edge pixel, and whether a pixel value of each adjacent pixel is greater than a first threshold, so as to expand the object backbone area in the second gray-scale image, which is extracted as a target object.

Abstract: A projection system including a first projection device and a first image capturing device is provided. The first projection device projects a first projection image. The first projection image includes a plurality of different color lights and has color blocks of different brightnesses formed by the plurality of different color lights. The first image capturing device captures the first projection image to generate a first captured image. The first image capturing device includes a first processor. The first processor converts the first captured image into a first converted image according to a first conversion matrix. A color gradation adjustment operation is performed on the first converted image to output an adjustment signal, and the first projection device adjusts the projected first projection image according to the adjustment signal. An image color correction method is also provided.

Abstract: A method and a system for inspecting a display image are provided. A test image is displayed on a display surface through a display device. An optical inspection image is generated by photographing the test image on the display surface by an image capturing device. A brightness channel image and a color channel image of the optical inspection image are obtained by performing color space transformation on the optical inspection image. Background estimation is applied to the color channel image to obtain a color channel background image. At least one color nonuniformity region in the test image is obtained by respectively comparing an inspection reference value with multiple color component pixel values of the color channel background image.

Abstract: An electronic device having a neural network framework for estimating optical flow is provided. The electronic device is connected to an image acquiring unit, which acquires images to be analyzed. The electronic device includes a storage unit, a feature extraction unit, an optical flow estimation unit and a refining unit. The storage unit stores a feature extraction module. The feature extraction unit is connected to the image acquiring unit and the storage unit. The optical flow estimation unit is connected to the feature extraction unit to generate an estimated optical flow. The refining unit is connected to the optical flow estimation unit to input the estimated optical flow to a refining module to obtain an estimated optical flow result. A method for estimating optical flow is also provided to reduce the number of training parameters required for estimating optical flow, thereby reducing a training time and improving training stability.

Abstract: A projection system including a first projection device and a first image capturing device is provided. The first projection device projects a first projection image. The first projection image includes a plurality of different color lights and has color blocks of different brightnesses formed by the plurality of different color lights. The first image capturing device captures the first projection image to generate a first captured image. The first image capturing device includes a first processor. The first processor converts the first captured image into a first converted image according to a first conversion matrix. A color gradation adjustment operation is performed on the first converted image to output an adjustment signal, and the first projection device adjusts the projected first projection image according to the adjustment signal. An image color correction method is also provided.

Abstract: A supervised training device and method for training a neural network, and a supervised classification method and device based on the neural network are provided. The training device includes a storage unit and a processor. The processor accesses and executes a sampling module, a labelling module and a training module in the storage unit. The sampling module samples a first image and a second image from a first dataset. The labelling module tags the first image in response to a first control instruction to generate a first tagged image, and generates label data according to the first tagged image and the second image. The training module trains the neural network according to the label data.