Target detection method and device

Abstract

The invention discloses a target detection method and device. The method comprises the steps: obtaining a plurality of bounding boxes on a target image and the scores of the bounding boxes, and enabling the scores to be used for representing the confidence of a target object contained in the bounding boxes; dividing the target image to obtain a plurality of grids, and determining the grids to which the plurality of bounding boxes belong; traversing the plurality of bounding boxes, calculating the overlapping degree of a reference bounding box and an adjacent bounding box, and obtaining a target bounding box according to the overlapping degree; and determining a target detection result according to the score of the target bounding box. According to the embodiment of the invention, grid division is carried out on the target image, the grid to which the bounding box belongs is determined, and then the adjacent relation of the bounding box is determined according to the adjacent relation between the grids, so that only the overlapping degree of the reference bounding box and the adjacent bounding box needs to be calculated during NMS algorithm operation, the time complexity of NMS is effectively reduced, and the calculation efficiency is improved. Target detection efficiency is improved.

Description

technical field [0001] The present application relates to the field of image processing, in particular to a target detection method and device. Background technique [0002] With the rapid development of deep learning and computer vision, related technologies have been widely used in many fields. Object Detection is an important part of image processing. Its task is to find out all the objects (objects) of interest in the image and determine their position and size. It is one of the core issues in the field of machine vision. [0003] When the deep learning method is used for target detection, it can include two steps. First, a bounding box (bbox) and a bounding box score are generated on the target image. The higher the score of the bounding box, the object in the bbox box is The probability of the target object is also greater; filter out the high-scoring bounding box whose bounding box score meets the preset conditions, and determine the object in the high-scoring boundi...

AI Technical Summary

Problems solved by technology

[0005] However, in the NMS algorithm, each selected high-scoring box has to perform IOU calculation with all low-scoring boxes with a score lower than itself. In the worst case, the algorithm complexity is O(N^2), so when the bbox box When the number increases, the time consumption will increase exponentially, and the increase in NMS time consumption will lead to a decrease in the detection frame rate, which will seriously affect the detection effect of the algorithm

Images