摘要:
Disclosed are an object detection method and an object detection device. The object detection method comprises a step of obtaining plural detection results of a current frame according to plural object detection methods; a step of setting initial probabilities of the plural detection results of the current frame; a step of calculating a movement frequency distribution diagram representing movement frequencies of respective pixels in the current frame; a step of obtaining detection results of a previous frame; a step of updating the probabilities of the plural detection results of the current frame; and a step of determining a final list of detected objects based on the updated probabilities of the plural detection results of the current frame.
摘要:
Disclosed are an object detection method and an object detection device. The object detection method comprises a step of obtaining plural detection results of a current frame according to plural object detection methods; a step of setting initial probabilities of the plural detection results of the current frame; a step of calculating a movement frequency distribution diagram representing movement frequencies of respective pixels in the current frame; a step of obtaining detection results of a previous frame; a step of updating the probabilities of the plural detection results of the current frame; and a step of determining a final list of detected objects based on the updated probabilities of the plural detection results of the current frame.
摘要:
Disclosed are a method and apparatus for recognizing a landmark in a panoramic image. The method includes steps of performing projection transformation on the panoramic image so as to generate a projection image; conducting semantic segmentation on the projection image so as to determine a landmark region and a road surface region; correcting distortion in the landmark region so as to produce a corrected landmark region; and recognizing the landmark in the corrected landmark region.
摘要:
Disclosed are an image matching method and apparatus. The image matching method is inclusive of steps of obtaining a panoramic image of at least one subspace in a 3D space and a 2D image of the 3D space; acquiring a 2D image of the at least one subspace in the 3D space; performing 3D reconstruction on the panoramic image of the at least one subspace, and procuring a projection image corresponding to the panoramic image of the at least one subspace; and attaining a matching relationship between the panoramic image of the at least one subspace and the 2D image of the at least one subspace, and establishing an association relationship between the panoramic image of the at least one subspace and the 2D image of the at least one subspace between which the matching relationship has been generated.
摘要:
A person counting method and a person counting system are provided. The method includes extracting a group of person images to obtain a first image set; dividing the first image set into first and second subsets based on whether a related image exists in a second image set, and reusing a person ID of the related image; estimating posture patterns of images in the first subset, and storing the images in the first subset into an image library based on person IDs and the posture patterns; and selecting a target image whose similarity to an image in the second subset is highest from the image library, reusing a person ID of the target image when the similarity is greater than a threshold, and assigning a new person ID and incrementing a person counter by 1 when the similarity is not greater than the threshold.
摘要:
Disclosed are a method and apparatus for recognizing a landmark in a panoramic image. The method includes steps of performing projection transformation on the panoramic image so as to generate a projection image; conducting semantic segmentation on the projection image so as to determine a landmark region and a road surface region; correcting distortion in the landmark region so as to produce a corrected landmark region; and recognizing the landmark in the corrected landmark region.
摘要:
An image processing method include projecting first and second images which are captured by two fisheye lenses onto first and second hemispherical projection models, respectively; determining a set of waiting-for-stitching surface region pairs on the first and second hemispherical projection models; extracting feature points from the two fisheye images and determining feature points matching each other to serving as pixels matching each other; determining a set of surface region pairs including the pixels matching each other on the first and second hemispherical projection models as the set of waiting-for-stitching surface region pairs; generating a customized projection model on the grounds of the set of waiting-for-stitching surface region pairs by letting a spherical distance between vertexes of each pair in the set of waiting-for-stitching surface region pairs be zero; and generating a third image by projecting the first and second images onto the customized projection model.
摘要:
The present invention provides a method and an apparatus for controlling a phase delay offset point of a modulator. The method comprises: acquiring backlight detection current signals output from a modulator in different working states, and determining harmonic amplitude values of the backlight detection current signals corresponding to the different working states; determining a detection value of a phase delay offset point corresponding to the modulator according to the determined harmonic amplitude values; comparing the detection value with a set target value of the phase delay offset point, and controlling a position of the phase delay offset point corresponding to the modulator according to the comparison result. The accuracy of controlling the position of the phase delay offset point of the modulator and the performance of the Differential Quadrature Phase Shift Keying (DQPSK) modulation system are improved through the technical solution.
摘要:
Method and apparatus for controlling bias point of DQPSK demodulator are disclosed. The method comprises: step 1: respectively applying first and second bias voltages to I-path and Q-path, and applying identical pilot voltage signals to I-path and Q-path (S202); step 2:executing filtering processing on I-path and Q-path differential current signals collected by balance receiver and determining θIand θQ (S204); step 3: performing feedback control to first and second bias voltages respectively according to θI and θQ so that θI and θQ respectively reaches expected bias point values of I-path and Q-path (S206); executing step 2 and 3 cyclically at preset regular intervals (S208), so that θI and θQ remains consistently the expected bias point values of I-path and Q-path. The solution enables bias point of DQPSK demodulator to be locked at any expected bias point value, facilitates realization of digitization, and is not easily influenced.
摘要:
Disclosed are a panoramic image based jump position determination method and apparatus. The method includes steps of obtaining first and second original panoramic images having the same horizontal width; performing, based on posture data of cameras respectively shooting the two original panoramic images, an adjustment process on at least one of the two original panoramic images, so as to acquire first and second panoramic images; after the adjustment process, conducting a feature extraction process regarding the two panoramic images, and choosing, based on a result of the feature extraction process, a plurality of pairs of image columns on the two panoramic images, so as to select at least one pair of first image columns from among the plurality of pairs of image columns; and determining, based on the at least one pair of first image columns, jump positions on the first and second panoramic images for jumping between the two.