Fast target detection method, electronic device, storage medium and system

Abstract

The invention provides a fast target detection method. The method comprises the steps of acquiring a training image in a preset database and a target image to be detected which needs to be detected; performing sample selection processing on the training image to obtain a positive sample image and a negative sample image; extracting the feature descriptors of the positive sample image and the negative sample image; inputting the feature descriptors into a word package model to obtain feature vectors; inputting the feature vectors into a nonlinear SVM classifier for training to obtain a trainedSVM classifier, and conducting simplification treatment on the trained classifier to obtain a simplified SVM classifier; extracting candidate region images and candidate region feature descriptors inthe target image to be tested; inputting the candidate region feature descriptors into the word package model, and obtaining candidate region feature vectors; according to the simplified SVM classifier candidate region feature vectors, detecting and obtaining a detection result. The invention provides the fast target detection method, and the training of the classifier through a large number of training images makes target detection no longer limited.

Description

technical field [0001] The invention relates to the field of computer vision, in particular to a fast target detection method, electronic equipment, storage media and a system. Background technique [0002] Object detection is a very important direction in the computer field. How to improve the real-time performance of the detector under the premise of ensuring the detection rate has always been a research hotspot in this direction. At present, support vector machines (i.e. SVM classifiers) are mostly used, but as the number of support vectors in the image to be detected increases, its classification speed also decreases. If the number of support vectors is too large, the classification speed of the SVM classifier will decrease. Very slow, which makes the SVM classifier very limited in occasions with high real-time requirements, and reduces the accuracy of target detection. Therefore, the existing target detection has a very slow detection speed, a certain limitation in de...

AI Technical Summary

Problems solved by technology

[0003] In order to overcome the deficiencies of the prior art, one of the purposes of the present invention is to provide a fast target detection method, which can solve the problems of slow detection speed and certain limitations in target detection.

[0004] The second object of the present invention is to provide an electronic device, which can solve the problems that the target detection speed is very slow and the detection has certain limitations.

[0005] The third object of the present invention is to provide a storage medium, which can solve the problem that the target detection speed is very slow and the detection has certain limitations.

Images