Github user imatiach-msft commented on a diff in the pull request:
https://github.com/apache/spark/pull/19439#discussion_r147737461
--- Diff: mllib/src/main/scala/org/apache/spark/ml/image/ImageSchema.scala
---
@@ -0,0 +1,252 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.spark.ml.image
+
+import java.awt.Color
+import java.awt.color.ColorSpace
+import java.io.ByteArrayInputStream
+import javax.imageio.ImageIO
+
+import org.apache.spark.annotation.{Experimental, Since}
+import org.apache.spark.input.PortableDataStream
+import org.apache.spark.sql.{DataFrame, Row, SparkSession}
+import org.apache.spark.sql.types._
+
+@Experimental
+@Since("2.3.0")
+object ImageSchema {
+
+ val undefinedImageType = "Undefined"
+
+ val imageFields = Array("origin", "height", "width", "nChannels",
"mode", "data")
+
+ val ocvTypes = Map(
+ undefinedImageType -> -1,
+ "CV_8U" -> 0, "CV_8UC1" -> 0, "CV_8UC3" -> 16, "CV_8UC4" -> 24
+ )
+
+ /**
+ * Schema for the image column: Row(String, Int, Int, Int, Int,
Array[Byte])
+ */
+ val columnSchema = StructType(
+ StructField(imageFields(0), StringType, true) ::
+ StructField(imageFields(1), IntegerType, false) ::
+ StructField(imageFields(2), IntegerType, false) ::
+ StructField(imageFields(3), IntegerType, false) ::
+ // OpenCV-compatible type: CV_8UC3 in most cases
+ StructField(imageFields(4), IntegerType, false) ::
+ // Bytes in OpenCV-compatible order: row-wise BGR in most cases
+ StructField(imageFields(5), BinaryType, false) :: Nil)
+
+ /**
+ * DataFrame with a single column of images named "image" (nullable)
+ */
+ val imageSchema = StructType(StructField("image", columnSchema, true) ::
Nil)
+
+ /**
+ * :: Experimental ::
+ * Gets the origin of the image
+ *
+ * @return The origin of the image
+ */
+ def getOrigin(row: Row): String = row.getString(0)
+
+ /**
+ * :: Experimental ::
+ * Gets the height of the image
+ *
+ * @return The height of the image
+ */
+ def getHeight(row: Row): Int = row.getInt(1)
+
+ /**
+ * :: Experimental ::
+ * Gets the width of the image
+ *
+ * @return The width of the image
+ */
+ def getWidth(row: Row): Int = row.getInt(2)
+
+ /**
+ * :: Experimental ::
+ * Gets the number of channels in the image
+ *
+ * @return The number of channels in the image
+ */
+ def getNChannels(row: Row): Int = row.getInt(3)
+
+ /**
+ * :: Experimental ::
+ * Gets the OpenCV representation as an int
+ *
+ * @return The OpenCV representation as an int
+ */
+ def getMode(row: Row): Int = row.getInt(4)
+
+ /**
+ * :: Experimental ::
+ * Gets the image data
+ *
+ * @return The image data
+ */
+ def getData(row: Row): Array[Byte] = row.getAs[Array[Byte]](5)
+
+ /**
+ * :: Experimental ::
+ * Check if the DataFrame column contains images (i.e. has ImageSchema)
+ *
+ * @param df DataFrame
+ * @param column Column name
+ * @return True if the given column matches the image schema
+ */
+ def isImageColumn(df: DataFrame, column: String): Boolean =
+ df.schema(column).dataType == columnSchema
--- End diff --
removed
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
