智能检测电击器算法模型的训练数据

本数据知识产权包含多角度、多场景下的电击器的X光安检图像，通过对图像的标注、抠图、增强、融合等处理，可作为优质样本训练生成智能检测电击器的算法模型，实现对藏匿在其他物品中或伪装成其他物品等复杂环境下的电击器精准识别。1、数据来源：应用X射线光源多角度、多场景下透射电击器，采集并建立其原始的X光数据图例库。 2、数据深度处理：对采集到的原始X光图像预标注坐标位置和品项类别，并对电击器的图像进行抠图处理。将抠出的图像与多场景图像分别进行几何变换、像素变换等增广处理。 3、检测模型生成算法规则：将处理后的电击器的X光安检图像和场景X光图像通过密度统计（像素值代表实物密度值）依据区域匹配原则进行融合，融合区域掩模作为数据标签与融合后的图像作为深度学习样本数据。还可通过调整抠图区域在场景图像区域的位置，获得不同的平均密度差值，训练生成可精准定位、精准识别电击器的智能检测模型。区域匹配原则按照Mask*(α*ρ抠图图像+β*ρ场景图像)，融合后的图像处理公式按照Mask*(α*ρ抠图图像+β*ρ抠图图像)+（1-Mask）*ρ场景图像。（所述公式中：Mask为图像掩膜，图像目标区域值为1，目标区域外值为2，ρ为密度值，α、β指系数）检测模型可对多场景下的电击器精准识别，同时将目标物的位置及所在X光图像信息记录标出。进一步的还可根据目标物位置信息推算目标物尺寸信息。

This intellectual property dataset contains X-ray security images of stun guns captured from multiple angles and scenarios. After processing such as image annotation, object matting, augmentation and fusion, it can be used as high-quality samples to train intelligent stun gun detection algorithm models, enabling accurate identification of stun guns hidden in other items or disguised as other items in complex environments.

1. Data Source: Use X-ray sources to perform transmission imaging of stun guns from multiple angles and scenarios, collect and establish an original X-ray image database.

2. In-depth Data Processing: Pre-annotate the coordinate positions and item categories for the collected original X-ray images, and conduct object matting processing on the stun gun images. Then, perform augmentation treatments including geometric transformations and pixel transformations on the matting-extracted images and multi-scenario images respectively.

3. Algorithm Rules for Detector Model Generation: Fuse the processed X-ray security images of stun guns and scenario X-ray images based on the regional matching principle via density statistics, where pixel values represent physical density values. The fusion region mask is used as the data label, and the fused image is used as deep learning sample data. Additionally, adjusting the position of the matting-extracted region within the scenario image can obtain different average density differences, which helps train an intelligent detection model that can accurately position and recognize stun guns.

The regional matching principle follows the formula: Mask*(α*ρ_matting_image + β*ρ_scenario_image), and the image processing formula for the fused result follows: Mask*(α*ρ_matting_image + β*ρ_matting_image) + (1-Mask)*ρ_scenario_image. (In the above formulas: Mask is the image mask, with a value of 1 in the target region of the image and 2 outside the target region; ρ represents density value; α and β are coefficients.)

The trained detection model can accurately identify stun guns in various scenarios, while recording and marking the position of the target object and the information of the X-ray image where it is located. Furthermore, the size information of the target object can be inferred based on its position information.

搜集汇总

数据集介绍

特点

该数据集包含1656条电击器的X光安检图像，数据来源于自行产生，更新频次为半年度。数据集通过多角度、多场景下的图像采集和处理，用于训练智能检测电击器的算法模型，能够在复杂环境下实现电击器的精准识别和定位。