Mike,
to solve these sorts of problems, you can try working in the opposite
direction.  When I want to de-serialize a document of a particular schema,
here's what I do:

1. infer the schema from the document using xsd.exe (only necessary if no
schema is available.  Really, the doc publisher should give you a schema,
you shouldn't have to infer it)

2. generate classes from that schema, using xsd.exe /c

3. build a test driver (20 lines of code) that instantiates the class and
then serializes it.  Compare to the xml document you are trying to match.
You can then iterate on your schema or the generated class, tweaking it to
match what you want.

-----

In your case,

A. the
xmlns:someobj=http://www.abcinc.com/objectdefinition
    is not valid XML.   Need quotes around this namespace value.

B. you did not specify the rest of the XML doc, but the assignment of the
someobj prefix corresponding to the http://www.abcinc.com/objectdefinition 
namespace seems to be unnecessary.   This will not break your
de-serialization, but it is superfluous.

C. the class I got by following the steps above was this:

[System.Xml.Serialization.XmlTypeAttribute(Namespace="http://www.abcinc.com/objectdefinition")]
[System.Xml.Serialization.XmlRootAttribute(Namespace="http://www.abcinc.com/objectdefinition",
IsNullable=false)]
public class SomeObject {
   public string SomeField;
}

And it de-serialized your XML (with the necessary quotes added, and one
element added for illustration) just fine.

D. to produce a similar XML doc by serializing, use
  XmlSerializerNamespaces ns = new XmlSerializerNamespaces();
  ns.Add( "", "http://www.abcinc.com/objectdefinition" );  // this will be
the default (no prefix) namespace in the XML doc, eg,  xmlns="..."

before serializing, and be sure to specify this ns in the Serialize() step.

  s1.Serialize(xmlWriter, so, ns);

-Dino