RADAR DATA PROCESSOR, OBJECT DISCRIMINATION DEVICE, RADAR DATA PROCESSING METHOD AND OBJECT DISCRIMINATION METHOD

• Main Authors: NAKAGAWA YOICHI, RYU IKETSU, YASUKI SHIN
• Format: Patent
• Language: English; Japanese
• Published: 30.01.2020
• Subjects: ANALOGOUS ARRANGEMENTS USING OTHER WAVES; CALCULATING; COMPUTING; COUNTING; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION ORRERADIATION OF RADIO WAVES; MEASURING; PHYSICS; RADIO DIRECTION-FINDING; RADIO NAVIGATION; TESTING

To perform object discrimination based on radar data with high accuracy by using a machine learning model.SOLUTION: Radar data including respective pieces of information of reflection intensity and speed in each azimuth and each distance, are acquired. A radar detection image storing respective pieces of information of reflection intensity, speed and distance corresponding to a position of a pixel, is generated in each of a plurality of channels of respective pixels, on the basis of the radar data. Then the radar detection image is input to a machine learning model for image recognition, which is learned to perform object discrimination, and an object discrimination result output from the machine learning model is acquired. Especially, an object region is detected from the radar data in the whole observation area, the radar data of the object region are extracted from the radar data in the whole observation area, and on the basis of the radar data of the object region, the radar detection image of the object region is generated.SELECTED DRAWING: Figure 2 【課題】機械学習モデルを用いることで、レーダデータに基づく物体判別を高精度に行うことができるようにする。【解決手段】各方位および各距離における反射強度および速度の各情報を含むレーダデータを取得し、そのレーダデータに基づいて、各画素の複数のチャンネルにそれぞれ、その画素の位置に対応する反射強度、速度および距離の各情報を格納したレーダ検出画像を生成し、そのレーダ検出画像を、物体判別を行うように学習した画像認識用の機械学習モデルに入力して、その機械学習モデルから出力される物体判別結果を取得する。特に、観測エリア全体のレーダデータから物体領域を検出して、観測エリア全体のレーダデータから物体領域のレーダデータを抽出して、その物体領域のレーダデータに基づいて、物体領域のレーダ検出画像を生成する。【選択図】図２