Method and device for detecting concealed objects in microwave images

#Embodiments of the present disclosure generally relate to image processing technology, and in particular relate to a method and device for detecting concealed objects in microwave images.

#Microwaves have a certain penetration during propagation. Through microwave imaging technology, it can obtain the image of the object under the clothes of the scanned (human) body, and then automatically complete the detection of the dangerous objects such as metal instruments, ceramic knives, unknown liquids, and powders which are under the clothes of the scanned body based on the image of the object. During the inspection process, if the detection of suspicious dangerous objects is performed by the directly visual observation of security operators, it will have huge consumption in manpower, financial resources and time. Therefore, it is of great significance to design an automatic detection method to perform the automatic detection of concealed dangerous objects with different attributes and types which are possible to appear on a body shown in a microwave image.

#The existing automatic object detection methods are generally divided into two types: the first one is for visible light image data; the second one is for microwave image data, while it is generally for single image detection. The above-mentioned two methods are not applicable to the security inspection system of microwave images for the main reasons are as follows: 1) the imaging mechanism of microwave images and visible light images are essentially different, and the dangerous objects detection method applied to visible light images cannot be directly applied to microwave images since the microwave image has low grayscale, low resolution, and is affected by speckle multiplicative noise; 2) since dangerous objects will show certain similarities and differences in adjacent images due to the difference in sampling angles, and it will depend on imaging quality too much and is easy to cause false detection when the detection is only performed on a single image, hence the feature detection method based on a single image cannot meet the requirement of the accuracy of the automatic detection of dangerous objects in a microwave security inspection system.

#FIG. 1 is a flow chart of a method for detecting concealed dangerous objects in microwave images according to an embodiment of the present disclosure.

#FIG. 2 is a schematic diagram of three obtained microwave images according to a specific embodiment of the present disclosure.

#FIG. 3 is a schematic diagram of a second dangerous-object area identified in a microwave image according to an embodiment of the present disclosure.

#In a microwave security inspection system (e.g., a near-field millimeter wave imaging security inspection system), a microwave image is formed after a microwave signal is transmitted, received, and the received signal is imaged, then an automatic detection process for the concealed dangerous objects is performed based on the microwave image information, and then the results of the automatic detection is presented to the system operator to assist the system operator in making determinations, thereby completing the security inspection process.

#Under normal circumstances, the interested dangerous objects are carious in their properties: in the aspect of materials, there are liquid, metal, plastic, powder, and the like; in the aspect of shape, there are mainly long strip, acute angle, short in width, and the like. In order to realize the automatic detection of the concealed dangerous objects in microwave images and achieve high detection accuracy, for the shown characteristics of the interested dangerous objects with different materials, shapes, sizes, and the like in microwave imaging results images, it is needed to analysis the content information of the microwave image, and the specific analysis results are as follows:

#For a single image, in the aspect of materials, the interested dangerous objects of the material such as metal, liquid, plastic, and powder appear in a microwave image as a partial area with a higher or lower grayscale value; in the aspect of shape, the objects of different shape such as long strip and acute angle appear in the imaging results as more obvious edge structure features. For multiple adjacent images at different angles, due to the difference in imaging angles, there is a certain difference between the grayscale value of the imaging of the same interested dangerous object and the body imaging grayscale value, which results in at different angles that the interested dangerous objects appear as having a higher/lower grayscale value or a grayscale value similar to a (human) body in the microwave image.