阅读说明：本技术 一种无人机检测图像识别方法 (A kind of unmanned plane detection image recognition methods ) 是由 朱姝 于 2018-05-21 设计创作，主要内容包括：本发明公开了一种无人机检测图像识别方法,涉及无人机检测图像识别,包括无人机检测类别样本、待识别图像,还包括以下步骤：S1,获取来自无人机检测检测采集点的待识别图像,在空间域中进行分解域变换,使图像信息分解到的多个尺度空间中；S2,利用空间域梯度算子进行待识别图像的边缘检测；S3,在图像变换域中,利用小波进行待识别图像的边缘检测；S4,提取空间域和变换域中边缘检测获取的待识别图像的特征向量；S5,利用人工神经网络进行模式识别。本发明能够有效提取无人机检测的边缘,作为理想的特征；能够依据空间域和变换域双重检测,提高检测精度。(The invention discloses a kind of unmanned plane detection image recognition methods, it is related to the identification of unmanned plane detection image, including unmanned machine testing classification sample, images to be recognized, it is further comprising the steps of: S1, obtain the images to be recognized from unmanned machine testing detection collection point, decomposition field transformation is carried out in the spatial domain, in the multiple scale spaces for decomposing image information；S2 carries out the edge detection of images to be recognized using spatial domain gradient operator；S3 carries out the edge detection of images to be recognized using small echo in image transform domain；S4 extracts the feature vector for the images to be recognized that edge detection obtains in spatial domain and transform domain；S5 carries out pattern-recognition using artificial neural network.The present invention can effectively extract the edge of unmanned machine testing, as ideal feature；Detection accuracy can be improved according to spatial domain and transform domain double check.)

1. a kind of unmanned plane detection image recognition methods, including unmanned machine testing classification sample, images to be recognized, feature exist In further comprising the steps of:

S1 obtains the images to be recognized from unmanned machine testing detection collection point, carries out decomposition field transformation in the spatial domain, make figure In the multiple scale spaces decomposed as information；

S2 carries out the edge detection of images to be recognized using spatial domain gradient operator；

S3 carries out the edge detection of images to be recognized using small echo in image transform domain；

S4 extracts the feature vector for the images to be recognized that edge detection obtains in spatial domain and transform domain；

S5, carries out pattern-recognition using artificial neural network, and unmanned machine testing classification sample is carried out off-line training, determines power Value carries out operation with the feature vector that S4 is obtained, realizes the identification of images to be recognized.

2. a kind of unmanned plane detection image recognition methods according to claim 1, which is characterized in that the decomposition in the S1 Domain transformation is converted using Multiscale Wavelet Decomposition domain.

3. a kind of unmanned plane detection image recognition methods according to claim 1, which is characterized in that the gradient in the S2 Operator uses Gauss-Laplace.

4. a kind of unmanned plane detection image recognition methods according to claim 1, which is characterized in that the edge in the S3 Testing result is recorded using chained list.

5. a kind of unmanned plane detection image recognition methods according to claim 1, which is characterized in that the feature in the S4 Vector uses statistical nature.

Technical field

The present invention relates to the identifications of unmanned plane detection image, and in particular to a kind of unmanned plane detection image recognition methods.

Background technique

UAV abbreviation unmanned plane, english abbreviation UAV are to utilize radio robot and the program provided for oneself The not manned aircraft of control device manipulation, or fully or intermittently automatically operated by car-mounted computer.With it is manned Aircraft is compared, and unmanned plane is often more suitable for those too slow-witted, dirty or dangerous tasks.Unmanned plane presses application field, can be divided into It is military with it is civilian.Military aspect, unmanned plane are divided into reconnaissance plane and target drone.Civilian aspect, unmanned plane+industrial application are unmanned planes Really just need；It taking photo by plane, agricultural, plant protection, miniature self-timer, express transportation, disaster relief, observation wild animal, monitoring at present The application in infectious disease, mapping, news report, electric inspection process, the disaster relief, movies-making, manufacture romance etc. field, is greatly expanded The purposes of unmanned plane itself, developed country also in actively extension industrial application and develop unmanned air vehicle technique.Therefore, to unmanned plane Quality testing, play increasingly important role in future society.At present there is identification error, exist in unmanned machine testing classification The problem of being difficult to when interference.

Summary of the invention

Exist the technical problem to be solved by the present invention is to unmanned machine testing classification at present and identifies that error, there are difficult when interference The problem of to identify, and it is an object of the present invention to provide a kind of unmanned plane detection image recognition methods, solves the above problems.

The present invention is achieved through the following technical solutions:

A kind of unmanned plane detection image recognition methods, including unmanned machine testing classification sample, images to be recognized, further include with Lower step:

S1 obtains the images to be recognized from unmanned machine testing detection collection point, carries out decomposition field transformation in the spatial domain, In the multiple scale spaces for decomposing image information；

S2 carries out the edge detection of images to be recognized using spatial domain gradient operator；

S3 carries out the edge detection of images to be recognized using small echo in image transform domain；

S4 extracts the feature vector for the images to be recognized that edge detection obtains in spatial domain and transform domain；

S5, carries out pattern-recognition using artificial neural network, and unmanned machine testing classification sample is carried out off-line training, is determined Weight carries out operation with the feature vector that S4 is obtained, realizes the identification of images to be recognized.

Further, the decomposition field transformation in the S1 is converted using Multiscale Wavelet Decomposition domain.

Further, the gradient operator in the S2 uses Gauss-Laplace.

Further, the edge detection results in the S3 are recorded using chained list.

Further, the feature vector in the S4 uses statistical nature.

Compared with prior art, the present invention having the following advantages and benefits:

1, a kind of unmanned plane detection image recognition methods of the present invention, can effectively extract the edge of unmanned machine testing, as Ideal feature；

2, a kind of unmanned plane detection image recognition methods of the present invention, can mention according to spatial domain and transform domain double check High measurement accuracy.

Specific embodiment

To make the objectives, technical solutions, and advantages of the present invention clearer, below with reference to embodiment, the present invention is made Further to be described in detail, exemplary embodiment of the invention and its explanation for explaining only the invention, are not intended as to this The restriction of invention.