To provide a passenger compartment authentication program and a passenger compartment authentication device with low facility cost.SOLUTION: First, in step 801, RGB color image data of all passenger compartments 10-1, 10-2, 10-3, and 10-4 are acquired from a camera 2. In step 802, grayscale image data of all passenger compartments 10-1, 10-2, 10-3, 10-4 are converted into black and white binarized data. In step 803, it is determined whether an initial state (S="1") or a non-initial state (S="0"). As a result, if it is in the initial state (S="1"), the process proceeds to steps 804 to 808, in which a quick vacancy determination is performed using the passenger compartment feature images of the passenger compartments 10-1, 10-2, 10-3, and 10-4, for example, the "P" mark image. If it is in the non-initial state (S="0"), the process proceeds to steps 809 and 810, in which highly accurate vacancy determination is performed by estimation processing using a well-learned learning control model 106 of each passenger compartments 10-1, 10-2, 10-3, and 10-4.SELECTED DRAWING: Figure 8
【課題】設備コストが低い車室認証プログラム及び車室認証装置を提供する。【解決手段】始めに、ステップ801にて、カメラ2より全車室10-1、10-2、10-3、10-4のRGBカラー画像データを取得する。ステップ802にて、全車室10-1、10-2、10-3、10-4のグレースケール画像データを黒白2値化データに変換する。ステップ803にて、初期状態(S=“1”)か非初期状態(S=“0”)かを判別する。この結果、初期状態(S=“1”)であれば、ステップ804~808に進み、各車室10-1、10-2、10-3、10-4の車室特徴画像たとえば「P」マーク画像による迅速な空満室判定を行う。非初期状態(S=“0”)であれば、ステップ809、810に進み、各車室10-1、10-2、10-3、10-4の十分学習した学習済の学習制御モデル106を用いた推定処理による高精度の空満室判定を行う。【選択図】 図8
PASSENGER COMPARTMENT AUTHENTICATION PROGRAM AND PASSENGER COMPARTMENT AUTHENTICATION DEVICE
車室認証プログラム及び車室認証装置
2022-04-14
Patent
Electronic Resource
Japanese
PASSENGER COMPARTMENT MONITORING METHOD AND PASSENGER COMPARTMENT MONITORING DEVICE
European Patent Office | 2021
|Passenger compartment of vehicle and vehicle with passenger compartment
European Patent Office | 2022
|